Professor Adrian Hilton FREng FIAPR FIET


Director Surrey Institute for People-Centred AI Director of Centre for Vision, Speech and Signal Processing
Fellow of the Royal Academy of Engineering; Fellow of the International Association for Pattern Recognition; Royal Society Wolfson Research Merit Award Holder in 4D Computer Vision; Distinguished Professor of the University of Surrey

About

Areas of specialism

Computer Vision; Machine Learning; 4D Vision; Computer Graphics; Virtual and Augmented Reality; Light Fields; Animation; Audio-Visual Signal Processing; 3D Computer Vision; Machine Perception ; AI; Digital Doubles

University roles and responsibilities

  • Director, Centre for Vision, Speech and Signal Processing (CVSSP)
  • Distinguished Professor of Computer Vision
  • Head of the Visual Media Research Lab. (V-Lab)

    Previous roles

    2008 - 2011
    Royal Society Industry Fellow with Framestore - Digital Doubles for Film Production
    https://www.framestore.com
    2014 - 2016
    Head of Department (Research), Department of Electrical and Electronic Engineering
    University of Surrey
    1997 - 2002
    EPSRC Advanced Research Fellow - Visual Reconstruction of Dynamic Scenes and People
    University of Surrey

    News

    Research

    Research interests

    Research projects

    Research collaborations

    Publications

    ARMIN MUSTAFA, Chris Russell, Adrian Hilton (2022)4D Temporally Coherent Multi-Person Semantic Reconstruction and Segmentation, In: International journal of computer vision Springer

    We introduce the first approach to solve the challenging problem of automatic 4D visual scene understanding for complex dynamic scenes with multiple interacting people from multi-view video. Our approach simultaneously estimates a detailed model that includes a per-pixel semantically and temporally coherent reconstruction, together with instance-level segmentation exploiting photo-consistency, semantic and motion information. We further leverage recent advances in 3D pose estimation to constrain the joint semantic instance segmentation and 4D temporally coherent reconstruction. This enables per person semantic instance segmentation of multiple interacting people in complex dynamic scenes. Extensive evaluation of the joint visual scene understanding framework against state-of-the-art methods on challenging indoor and outdoor sequences demonstrates a significant (≈40%) improvement in semantic segmentation, reconstruction and scene flow accuracy. In addition to the evaluation on several indoor and outdoor scenes, the proposed joint 4D scene understanding framework is applied to challenging outdoor sports scenes in the wild captured with manually operated wide-baseline broadcast cameras.

    Matthew James Bailey, Adrian Douglas Mark Hilton, Jean-Yves Guillemaut (2022)Finite Aperture Stereo: 3D Reconstruction of Macro-Scale Scenes, In: Proceedings of the 2021 IEEE/CVF International Conference on Computer Vision Workshops (ICCVW 2021)pp. 2474-2484 Institute of Electrical and Electronics Engineers (IEEE)

    While the accuracy of multi-view stereo (MVS) has continued to advance, its performance reconstructing challenging scenes from images with a limited depth of field is generally poor. Typical implementations assume a pinhole camera model, and therefore treat defocused regions as a source of outlier. In this paper, we address these limitations by instead modelling the camera as a thick lens. Doing so allows us to exploit the complementary nature of stereo and defocus information, and overcome constraints imposed by traditional MVS methods. Using our novel reconstruction framework, we recover complete 3D models of complex macro-scale scenes. Our approach demonstrates robustness to view-dependent materials, and outperforms state-of-the-art MVS and depth from defocus across a range of real and synthetic datasets.

    Hansung Kim, L Remaggi, Samuel Fowler, Philip JB Jackson, Adrian Hilton (2020)Acoustic Room Modelling Using 360 Stereo Cameras, In: IEEE Transactions on Multimedia23pp. 4117-4130 Institute of Electrical and Electronics Engineers (IEEE)

    In this paper we propose a pipeline for estimating acoustic 3D room structure with geometry and attribute prediction using spherical 360° cameras. Instead of setting microphone arrays with loudspeakers to measure acoustic parameters for specific rooms, a simple and practical single-shot capture of the scene using a stereo pair of 360° cameras can be used to simulate those acoustic parameters. We assume that the room and objects can be represented as cuboids aligned to the main axes of the room coordinate (Manhattan world). The scene is captured as a stereo pair using off-the-shelf consumer spherical 360 cameras. A cuboid-based 3D room geometry model is estimated by correspondence matching between captured images and semantic labelling using a convolutional neural network (SegNet). The estimated geometry is used to produce frequency-dependent acoustic predictions of the scene. This is, to our knowledge, the first attempt in the literature to use visual geometry estimation and object classification algorithms to predict acoustic properties. Results are compared to measurements through calculated reverberant spatial audio object parameters used for reverberation reproduction customized to the given loudspeaker set up.

    HANSUNG KIM, L Remaggi, Aloisio Dourado, T Campos, PHILIP J B JACKSON, Adrian Hilton (2021)Immersive audio-visual scene reproduction using semantic scene reconstruction from 360 cameras, In: Virtual reality : the journal of the Virtual Reality Society Springer

    AbstractAs personalised immersive display systems have been intensely explored in virtual reality (VR), plausible 3D audio corresponding to the visual content is required to provide more realistic experiences to users. It is well known that spatial audio synchronised with visual information improves a sense of immersion but limited research progress has been achieved in immersive audio-visual content production and reproduction. In this paper, we propose an end-to-end pipeline to simultaneously reconstruct 3D geometry and acoustic properties of the environment from a pair of omnidirectional panoramic images. A semantic scene reconstruction and completion method using a deep convolutional neural network is proposed to estimate the complete semantic scene geometry in order to adapt spatial audio reproduction to the scene. Experiments provide objective and subjective evaluations of the proposed pipeline for plausible audio-visual VR reproduction of real scenes.

    Hansung Kim, Luca Remaggi, Philip J. B. Jackson, Adrian Hilton (2020)Immersive Virtual Reality Audio Rendering Adapted to the Listener and the Room, In: Real VR – Immersive Digital Realitypp. 293-318 Springer Nature

    The visual and auditory modalities are the most important stimuli for humans. In order to maximise the sense of immersion in VR environments, a plausible spatial audio reproduction synchronised with visual information is essential. However, measuring acoustic properties of an environment using audio equipment is a complicated process. In this chapter, we introduce a simple and efficient system to estimate room acoustic for plausible spatial audio rendering using 360 ∘ cameras for real scene reproduction in VR. A simplified 3D semantic model of the scene is estimated from captured images using computer vision algorithms and convolutional neural network (CNN). Spatially synchronised audio is reproduced based on the estimated geometric and acoustic properties in the scene. The reconstructed scenes are rendered with synthesised spatial audio.

    Matthew James Bailey, Adrian Douglas Mark Hilton, Jean-Yves Guillemaut (2022)Finite Aperture Stereo, In: International journal of computer vision130pp. 2858-2884 Springer Nature

    Multi-view stereo remains a popular choice when recovering 3D geometry, despite performance varying dramatically according to the scene content. Moreover, typical pinhole camera assumptions fail in the presence of shallow depth of field inherent to macro-scale scenes; limiting application to larger scenes with diffuse reflectance. However, the presence of defocus blur can itself be considered a useful reconstruction cue, particularly in the presence of view-dependent materials. With this in mind, we explore the complimentary nature of stereo and defocus cues in the context of multi-view 3D reconstruction; and propose a complete pipeline for scene modelling from a finite aperature camera that encompasses image formation, camera calibration and reconstruction stages. As part of our evaluation, an ablation study reveals how each cue contributes to the higher performance observed over a range of complex materials and geometries. Though of lesser concern with large apertures, the effects of image noise are also considered. By introducing pre-trained deep feature extraction into our cost function, we show a step improvement over per-pixel comparisons; as well as verify the cross-domain applicability of networks using largely in-focus training data applied to defocused images. Finally, we compare to a number of modern multi-view stereo methods, and demonstrate how the use of both cues leads to a significant increase in performance across several synthetic and real datasets.

    MATTHEW JAMES BAILEY, Adrian Hilton, Jean-Yves Guillemaut Finite Aperture Stereo Datasets CVSSP

    This landing page contains the datasets presented in the paper "Finite Aperture Stereo". The datasets are intended for defocus-based 3D reconstruction and analysis. Each download link contains images of a static scene, captured from multiple viewpoints and with different focus settings. The captured objects exhibit a range of reflectance properties and are physically small in scale. Calibration images are also available. A CC BY-NC licence is in effect. Use of this data must be for non-commercial research purposes. Acknowledgement must be given to the original authors by referencing the dataset DOI, the dataset web address, and the aforementioned publication. Re-distribution of this data is prohibited. Before downloading, you must agree with these conditions as presented on the dataset webpage.

    P Csakany, F Vajda, A Hilton (2006)Model Refinement by Iterative Normal-From-Shading, In: IET European Conference on Visual Media Productionpp. 181-181
    Hansung Kim, RJ Hughes, Luca Remaggi, Philip Jackson, Adrian Hilton, TJ Cox, B Shirley (2017)Acoustic Room Modelling using a Spherical Camera for Reverberant Spatial Audio Objects, In: Proceedings of the Audio Engineering Society142

    The ability to predict the acoustics of a room without acoustical measurements is a useful capability. The motivation here stems from spatial audio reproduction, where knowledge of the acoustics of a space could allow for more accurate reproduction of a captured environment, or for reproduction room compensation techniques to be applied. A cuboid-based room geometry estimation method using a spherical camera is proposed, assuming a room and objects inside can be represented as cuboids aligned to the main axes of the coordinate system. The estimated geometry is used to produce frequency-dependent acoustic predictions based on geometrical room modelling techniques. Results are compared to measurements through calculated reverberant spatial audio object parameters used for reverberation reproduction customized to the given loudspeaker set up.

    J Edge, A Hilton, PHILIP J B JACKSON (2008)Parameterisation of Speech Lip Movements, In: Proceedings of International Conference on Auditory-visual Speech Processing

    In this paper we describe a parameterisation of lip movements which maintains the dynamic structure inherent in the task of producing speech sounds. A stereo capture system is used to reconstruct 3D models of a speaker producing sentences from the TIMIT corpus. This data is mapped into a space which maintains the relationships between samples and their temporal derivatives. By incorporating dynamic information within the parameterisation of lip movements we can model the cyclical structure, as well as the causal nature of speech movements as described by an underlying visual speech manifold. It is believed that such a structure will be appropriate to various areas of speech modeling, in particular the synthesis of speech lip movements.

    Jon Francombe, James Woodcock, Richard J. Hughes, Russell Mason, Andreas Franck, Chris Pike, Tim Brookes, William J. Davies, Philip J.B. Jackson, Trevor J. Cox, Filippo M. Fazi, Adrian Hilton (2018)Qualitative evaluation of media device orchestration for immersive spatial audio reproduction, In: Journal of the Audio Engineering Society66(6)pp. 414-429 Audio Engineering Society

    The challenge of installing and setting up dedicated spatial audio systems can make it difficult to deliver immersive listening experiences to the general public. However, the proliferation of smart mobile devices and the rise of the Internet of Things mean that there are increasing numbers of connected devices capable of producing audio in the home. ____Media device orchestration" (MDO) is the concept of utilizing an ad hoc set of devices to deliver or augment a media experience. In this paper, the concept is evaluated by implementing MDO for augmented spatial audio reproduction using objectbased audio with semantic metadata. A thematic analysis of positive and negative listener comments about the system revealed three main categories of response: perceptual, technical, and content-dependent aspects. MDO performed particularly well in terms of immersion/envelopment, but the quality of listening experience was partly dependent on loudspeaker quality and listener position. Suggestions for further development based on these categories are given.

    Andrew Gilbert, Matthew Trumble, Charles Malleson, Adrian Hilton, John Collomosse (2018)Fusing Visual and Inertial Sensors with Semantics for 3D Human Pose Estimation, In: International Journal of Computer Vision Springer Verlag

    We propose an approach to accurately esti- mate 3D human pose by fusing multi-viewpoint video (MVV) with inertial measurement unit (IMU) sensor data, without optical markers, a complex hardware setup or a full body model. Uniquely we use a multi-channel 3D convolutional neural network to learn a pose em- bedding from visual occupancy and semantic 2D pose estimates from the MVV in a discretised volumetric probabilistic visual hull (PVH). The learnt pose stream is concurrently processed with a forward kinematic solve of the IMU data and a temporal model (LSTM) exploits the rich spatial and temporal long range dependencies among the solved joints, the two streams are then fused in a final fully connected layer. The two complemen- tary data sources allow for ambiguities to be resolved within each sensor modality, yielding improved accu- racy over prior methods. Extensive evaluation is per- formed with state of the art performance reported on the popular Human 3.6M dataset [26], the newly re- leased TotalCapture dataset and a challenging set of outdoor videos TotalCaptureOutdoor. We release the new hybrid MVV dataset (TotalCapture) comprising of multi- viewpoint video, IMU and accurate 3D skele- tal joint ground truth derived from a commercial mo- tion capture system. The dataset is available online at http://cvssp.org/data/totalcapture/.

    Jean-Yves Guillemaut, J Kilner, J Starck, Adrian Hilton (2007)Dynamic feathering: Minimising blending artefacts in view-dependent rendering, In: IET Conference Publications534(534 CP)

    Conventional view-dependent texture mapping techniques produce composite images by blending subsets of input images, weighted according to their relative influence at the rendering viewpoint, over regions where the views overlap. Geometric or camera calibration errors often result in a los s of detail due to blurring or double exposure artefacts which tends to be exacerbated by the number of blending views considered. We propose a novel view-dependent rendering technique which optimises the blend region dynamically at rendering time, and reduces the adverse effects of camera calibration or geometric errors otherwise observed. The technique has been successfully integrated in a rendering pipeline which operates at interactive frame rates. Improvement over state-of-the-art view-dependent texture mapping techniques are illustrated on a synthetic scene as well as real imagery of a large scale outdoor scene where large camera calibration and geometric errors are present.

    We propose a robust 3D feature description and registration method for 3D models reconstructed from various sensor devices. General 3D feature detectors and descriptors generally show low distinctiveness and repeatability for matching between different data modalities due to differences in noise and errors in geometry. The proposed method considers not only local 3D points but also neighbouring 3D keypoints to improve keypoint matching. The proposed method is tested on various multi-modal datasets including LIDAR scans, multiple photos, spherical images and RGBD videos to evaluate the performance against existing methods.

    Graham Miller, Adrian Hilton, J Starck (2005)Interactive Free-viewpoint Video, In: IEEE European Conf. on Visual Media Productionpp. 50-59
    J Starck, Graham Miller, Adrian Hilton (2006)Volumetric stereo with silhouette and feature constraints, In: British Machine Vision Conferencepp. 1189-1198

    This paper presents a novel volumetric reconstruction technique that combines shape-from-silhouette with stereo photo-consistency in a global optimisation that enforces feature constraints across multiple views. Human shape reconstruction is considered where extended regions of uniform appearance, complex self-occlusions and sparse feature cues represent a challenging problem for conventional reconstruction techniques. A unified approach is introduced to first reconstruct the occluding contours and left-right consistent edge contours in a scene and then incorporate these contour constraints in a global surface optimisation using graph-cuts. The proposed technique maximises photo-consistency on the surface, while satisfying silhouette constraints to provide shape in the presence of uniform surface appearance and edge feature constraints to align key image features across views.

    P Huang, J Starck, A Hilton (2007)A Study of Shape Similarity for Temporal Surface Sequences of People, In: IEEE Int.Conf. on 3D Imaging and Modelingpp. 408-418
    Armin Mustafa, Hansung Kim, Jean-Yves Guillemaut, Adrian Hilton (2015)General Dynamic Scene Reconstruction from Multiple View Video, In: 2015 IEEE International Conference on Computer Vision (ICCV)pp. 900-908 IEEE

    This paper introduces a general approach to dynamic scene reconstruction from multiple moving cameras without prior knowledge or limiting constraints on the scene structure, appearance, or illumination. Existing techniques for dynamic scene reconstruction from multiple wide-baseline camera views primarily focus on accurate reconstruction in controlled environments, where the cameras are fixed and calibrated and background is known. These approaches are not robust for general dynamic scenes captured with sparse moving cameras. Previous approaches for outdoor dynamic scene reconstruction assume prior knowledge of the static background appearance and structure. The primary contributions of this paper are twofold: an automatic method for initial coarse dynamic scene segmentation and reconstruction without prior knowledge of background appearance or structure; and a general robust approach for joint segmentation refinement and dense reconstruction of dynamic scenes from multiple wide-baseline static or moving cameras. Evaluation is performed on a variety of indoor and outdoor scenes with cluttered backgrounds and multiple dynamic non-rigid objects such as people. Comparison with state-of-the-art approaches demonstrates improved accuracy in both multiple view segmentation and dense reconstruction. The proposed approach also eliminates the requirement for prior knowledge of scene structure and appearance.

    JJ Kilner, J Starck, A Hilton, JY Guillemaut, O Grau (2007)Dual Mode Deformable Models for Free-Viewpoint Video of Outdoor Sports Events, In: IEEE Int.Conf. on 3D Imaging and Modelingpp. 177-184
    AJ Stoddart, A Hilton (1996)Registration of multiple point sets, In: ICPR
    J Kittler, A Hilton, M Hamouz, J Illingworth (2006)3D Assisted Face Recognition: A Survey of 3D Imaging, Modelling and Recognition Approaches, In: CVPR
    P Huang, A Hilton, J Starck (2010)Shape Similarity for 3D Video Sequences of People, In: International Journal of Computer Vision89(2-3)pp. 362-381 Springer

    This paper presents a performance evaluation of shape similarity metrics for 3D video sequences of people with unknown temporal correspondence. Performance of similarity measures is compared by evaluating Receiver Operator Characteristics for classification against ground-truth for a comprehensive database of synthetic 3D video sequences comprising animations of fourteen people performing twentyeight motions. Static shape similarity metrics shape distribution, spin image, shape histogram and spherical harmonics are evaluated using optimal parameter settings for each approach. Shape histograms with volume sampling are found to consistently give the best performance for different people and motions. Static shape similarity is extended over time to eliminate the temporal ambiguity. Time-filtering of the static shape similarity together with two novel shape-flow descriptors are evaluated against temporal ground-truth. This evaluation demonstrates that shape-flow with a multi-frame alignment of motion sequences achieves the best performance, is stable for different people and motions, and overcome the ambiguity in static shape similarity. Time-filtering of the static shape histogram similarity measure with a fixed window size achieves marginally lower performance for linear motions with the same computational cost as static shape descriptors. Performance of the temporal shape descriptors is validated for real 3D video sequence of nine actors performing a variety of movements. Time-filtered shape histograms are shown to reliably identify frames from 3D video sequences with similar shape and motion for people with loose clothing and complex motion.

    O Grau, Adrian Hilton, J Kilner, Graham Miller, T Sargeant, J Starck (2006)A Free-Viewpoint Video System for Visualisation of Sports Scenes, In: International Broadcast ConventionSeptem
    JJ Kilner, J-Y Guillemaut, A Hilton (2009)Summarised Hierarchical Markov Models for Speed Invariant Action Matching., In: ICCV Workshop on Tracking Humans for the Evaluation of their Motion in Image Sequencespp. 1065-1072

    Action matching, where a recorded sequence is matched against, and synchronised with, a suitable proxy from a library of animations, is a technique for generating a synthetic representation of a recorded human activity. This proxy can then be used to represent the action in a virtual environment or as a prior on further processing of the sequence. In this paper we present a novel technique for performing action matching in outdoor sports environments. Outdoor sports broadcasts are typically multi-camera environments and as such reconstruction techniques can be applied to the footage to generate a 3D model of the scene. However due to poor calibration and matting this reconstruction is of a very low quality. Our technique matches the 3D reconstruction sequence against a predefined library of actions to select an appropriate high quality synthetic representation. A hierarchical Markov model combined with 3D summarisation of the data allows a large number of different actions to be matched successfully to the sequence in a rate-invariant manner without prior segmentation of the sequence into discrete units. The technique is applied to data captured at rugby and soccer games.

    Yang Liu, Adrian Hilton, J Chambers, Y Zhao, Wenwu Wang (2018)Non-zero diffusion particle flow SMC-PHD filter for audio-visual multi-speaker tracking, In: Proceedings of ICASSP 2018 IEEE

    The sequential Monte Carlo probability hypothesis density (SMC-PHD) filter has been shown to be promising for audio-visual multi-speaker tracking. Recently, the zero diffusion particle flow (ZPF) has been used to mitigate the weight degeneracy problem in the SMC-PHD filter. However, this leads to a substantial increase in the computational cost due to the migration of particles from prior to posterior distribution with a partial differential equation. This paper proposes an alternative method based on the non-zero diffusion particle flow (NPF) to adjust the particle states by fitting the particle distribution with the posterior probability density using the nonzero diffusion. This property allows efficient computation of the migration of particles. Results from the AV16.3 dataset demonstrate that we can significantly mitigate the weight degeneracy problem with a smaller computational cost as compared with the ZPF based SMC-PHD filter.

    C Budd, A Hilton (2009)Skeleton Driven Volumetric Deformation, In: ACM Symposium on Computer Animation
    A Hilton, J Goncalves (1995)3D Scene Representation Using a Deformable Surface, In: IEEE Workshop on Physics Based Modelling
    AJ Stoddart, S Lemke, A Hilton, T Renn (1996)Uncertainty estimation for surface registration, In: BMVC
    Luca Remaggi, Hansung Kim, Philip J. B. Jackson, Adrian Hilton (2019)Reproducing Real World Acoustics in Virtual Reality Using Spherical Cameras, In: Proceedings of the 2019 AES International Conference on Immersive and Interactive Audio Audio Engineering Society

    Virtual Reality (VR) systems have been intensely explored, with several research communities investigating the different modalities involved. Regarding the audio modality, one of the main issues is the generation of sound that is perceptually coherent with the visual reproduction. Here, we propose a pipeline for creating plausible interactive reverb using visual information: first, we characterize real environment acoustics given a pair of spherical cameras; then, we reproduce reverberant spatial sound, by using the estimated acoustics, within a VR scene. The evaluation is made by extracting the room impulse responses (RIRs) of four virtually rendered rooms. Results show agreement, in terms of objective metrics, between the synthesized acoustics and the ones calculated from RIRs recorded within the respective real rooms.

    J-Y Guillemaut, J Kilner, A Hilton (2009)Robust Graph-Cut Scene Segmentation and Reconstruction for Free-Viewpoint Video of Complex Dynamic Scenes, In: 2009 IEEE 12TH INTERNATIONAL CONFERENCE ON COMPUTER VISION (ICCV)pp. 809-816
    H Kim, A Hilton (2009)Graph-based Foreground Extraction in Extended Colour Space, In: Int.Conf.Image Processing (ICIP)
    Eng-Jon Ong, Adrian Hilton (2006)Learnt Inverse Kinematics for Animation Synthesis, In: Graphical Models685-6pp. 472-483 Elsevier

    Existing work on animation synthesis can be roughly split into two approaches, those that combine segments of motion capture data, and those that perform inverse kinematics. In this paper, we present a method for performing animation synthesis of an articulated object (e.g. human body and a dog) from a minimal set of body joint positions, following the approach of inverse kinematics. We tackle this problem from a learning perspective. Firstly, we address the need for knowledge on the physical constraints of the articulated body, so as to avoid the generation of a physically impossible poses. A common solution is to heuristically specify the kinematic constraints for the skeleton model. In this paper however, the physical constraints of the articulated body are represented using a hierarchical cluster model learnt from a motion capture database. Additionally, we shall show that the learnt model automatically captures the correlation between different joints through the simultaneous modelling their angles. We then show how this model can be utilised to perform inverse kinematics in a simple and efficient manner. Crucially, we describe how IK is carried out from a minimal set of end-effector positions. Following this, we show how this "learnt inverse kinematics" framework can be used to perform animation syntheses of different types of articulated structures. To this end, the results presented include the retargeting of a at surface walking animation to various uneven terrains to demonstrate the synthesis of a full human body motion from the positions of only the hands, feet and torso. Additionally, we show how the same method can be applied to the animation synthesis of a dog using only its feet and torso positions.

    C Malleson, M Klaudiny, A Hilton, J-Y Guillemaut (2013)Single-view RGBD-based reconstruction of dynamic human geometry, In: Proceedings of the IEEE International Conference on Computer Vision - Workshop on Dynamic Shape Capture and Analysis (4DMOD 2013)pp. 307-314

    We present a method for reconstructing the geometry and appearance of indoor scenes containing dynamic human subjects using a single (optionally moving) RGBD sensor. We introduce a framework for building a representation of the articulated scene geometry as a set of piecewise rigid parts which are tracked and accumulated over time using moving voxel grids containing a signed distance representation. Data association of noisy depth measurements with body parts is achieved by online training of a prior shape model for the specific subject. A novel frame-to-frame model registration is introduced which combines iterative closest-point with additional correspondences from optical flow and prior pose constraints from noisy skeletal tracking data. We quantitatively evaluate the reconstruction and tracking performance of the approach using a synthetic animated scene. We demonstrate that the approach is capable of reconstructing mid-resolution surface models of people from low-resolution noisy data acquired from a consumer RGBD camera. © 2013 IEEE.

    Y Li, A Hilton, J Illingworth (2002)A Relaxation Algorithm for Real-time Multiview 3D-Tracking, In: Image and Vision Computing2012pp. 841-859 Elsevier
    Armin Mustafa, Hansung Kim, Adrian Hilton (2016)4D Match Trees for Non-rigid Surface Alignment, In: . Springer (eds.), Computer Vision – ECCV 2016 14th European Conference, Amsterdam, The Netherlands, October 11–14, 2016, Proceedings, Part I9905(1)pp. 213-229

    This paper presents a method for dense 4D temporal alignment of partial reconstructions of non-rigid surfaces observed from single or multiple moving cameras of complex scenes. 4D Match Trees are introduced for robust global alignment of non-rigid shape based on the similarity between images across sequences and views. Wide-timeframe sparse correspondence between arbitrary pairs of images is established using a segmentation-based feature detector (SFD) which is demonstrated to give improved matching of non-rigid shape. Sparse SFD correspondence allows the similarity between any pair of image frames to be estimated for moving cameras and multiple views. This enables the 4D Match Tree to be constructed which minimises the observed change in non-rigid shape for global alignment across all images. Dense 4D temporal correspondence across all frames is then estimated by traversing the 4D Match tree using optical flow initialised from the sparse feature matches. The approach is evaluated on single and multiple view images sequences for alignment of partial surface reconstructions of dynamic objects in complex indoor and outdoor scenes to obtain a temporally consistent 4D representation. Comparison to previous 2D and 3D scene flow demonstrates that 4D Match Trees achieve reduced errors due to drift and improved robustness to large non-rigid deformations.

    Andrew Gilbert, Adrian Hilton, John Collomosse (2020)Semantic Estimation of 3D Body Shape and Pose using Minimal Cameras, In: 31st British Machine Vision Conference

    We aim to simultaneously estimate the 3D articulated pose and high fidelity volumetric occupancy of human performance, from multiple viewpoint video (MVV) with as few as two views. We use a multi-channel symmetric 3D convolutional encoder-decoder with a dual loss to enforce the learning of a latent embedding that enables inference of skeletal joint positions and a volumetric reconstruction of the performance. The inference is regularised via a prior learned over a dataset of view-ablated multi-view video footage of a wide range of subjects and actions, and show this to generalise well across unseen subjects and actions. We demonstrate improved reconstruction accuracy and lower pose estimation error relative to prior work on two MVV performance capture datasets: Human 3.6M and TotalCapture.

    Sam Fowler, Hansung Kim, Adrian Hilton (2018)Human-Centric Scene Understanding from Single View 360 Video, In: 2018 International Conference on 3D Vision (3DV)pp. 334-342 Institute of Electrical and Electronics Engineers (IEEE)

    In this paper, we propose an approach to indoor scene understanding from observation of people in single view spherical video. As input, our approach takes a centrally located spherical video capture of an indoor scene, estimating the 3D localisation of human actions performed throughout the long term capture. The central contribution of this work is a deep convolutional encoder-decoder network trained on a synthetic dataset to reconstruct regions of affordance from captured human activity. The predicted affordance segmentation is then applied to compose a reconstruction of the complete 3D scene, integrating the affordance segmentation into 3D space. The mapping learnt between human activity and affordance segmentation demonstrates that omnidirectional observation of human activity can be applied to scene understanding tasks such as 3D reconstruction. We show that our approach using only observation of people performs well against previous approaches, allowing reconstruction of occluded regions and labelling of scene affordances.

    We describe a novel framework for segmenting a time- and view-coherent foreground matte sequence from synchronised multiple view video. We construct a Markov Random Field (MRF) comprising links between superpixels corresponded across views, and links between superpixels and their constituent pixels. Texture, colour and disparity cues are incorporated to model foreground appearance. We solve using a multi-resolution iterative approach enabling an eight view high definition (HD) frame to be processed in less than a minute. Furthermore we incorporate a temporal diffusion process introducing a prior on the MRF using information propagated from previous frames, and a facility for optional user correction. The result is a set of temporally coherent mattes that are solved for simultaneously across views for each frame, exploiting similarities across views and time.

    A Hilton, J Illingworth (1997)Multi-Resolution Geometric Fusion, In: International Conference on Recent Advances in 3D Digital Imaging and Modeling
    C Budd, A Hilton (2009)Skeleton Driven Volumetric Laplacian Deformation, In: European Conference on Visual Media Production

    Typical colour digital cameras have a single sensor with a colour filter array (CFA), each pixel capturing a single channel (red, green or blue). A full RGB colour output image is generated by demosaicing (DM), i.e. interpolating to infer the two unobserved channels for each pixel. The DM approach used can have a significant effect on the quality of the output image, particularly in the presence of common imaging artifacts such as chromatic aberration (CA). Small differences in the focal length for each channel (lateral CA) and the inability of the lens to bring all three channels simultaneously into focus (longitudinal CA) can cause objectionable colour fringing artifacts in edge regions. These artifacts can be particularly severe when using low-cost lenses. We propose to use a set of simple neural networks to learn to jointly perform DM and CA correction, producing high quality colour images subject to severe CA as well as image noise. The proposed neural network-based joint DM and CA correction produces a significant improvement in image quality metrics (PSNR and SSIM) compared the baseline edge-directed linear interpolation approach preserving image detail and reducing objectionable false colour and comb artifacts. The approach can be applied in the production of high quality images and video from machine vision cameras with low cost lenses, thus extending the viability of such hardware to visual media production.

    Armin Mustafa, Hansung Kim, Adrian Hilton (2018)MSFD: Multi-scale segmentation based feature detection for wide-baseline scene reconstruction, In: IEEE Transactions on Image Processing28(3)pp. 1118-1132 Institute of Electrical and Electronics Engineers (IEEE)

    A common problem in wide-baseline matching is the sparse and non-uniform distribution of correspondences when using conventional detectors such as SIFT, SURF, FAST, A-KAZE and MSER. In this paper we introduce a novel segmentation based feature detector (SFD) that produces an increased number of accurate features for wide-baseline matching. A multi-scale SFD is proposed using bilateral image decomposition to produce a large number of scale-invariant features for wide-baseline reconstruction. All input images are over-segmented into regions using any existing segmentation technique like Watershed, Mean-shift, and SLIC. Feature points are then detected at the intersection of the boundaries of three or more regions. The detected feature points are local maxima of the image function. The key advantage of feature detection based on segmentation is that it does not require global threshold setting and can therefore detect features throughout the image. A comprehensive evaluation demonstrates that SFD gives an increased number of features which are accurately localised and matched between wide-baseline camera views; the number of features for a given matching error increases by a factor of 3-5 compared to SIFT; feature detection and matching performance is maintained with increasing baseline between views; multi-scale SFD improves matching performance at varying scales. Application of SFD to sparse multi-view wide-baseline reconstruction demonstrates a factor of ten increase in the number of reconstructed points with improved scene coverage compared to SIFT/MSER/A-KAZE. Evaluation against ground-truth shows that SFD produces an increased number of wide-baseline matches with reduced error.

    J-Y Guillemaut, J Kilner, A Hilton (2009)Robust Graph-Cut Scene Segmentation and Reconstruction for Free-Viewpoint Video of Complex Dynamic Scenes, In: IEEE Int.Conf. on Computer Vision, ICCVpp. 809-816
    J Starck, A Hilton (2005)Virtual View Synthesis of People from Multiple View Video, In: Graphical Models67(6)pp. 600-620 Elsevier

    This paper addresses the synthesis of virtual views of people from multiple view image sequences. We consider the target area of the multiple camera “3D Virtual Studio” with the ultimate goal of capturing video-realistic dynamic human appearance. A mesh based reconstruction framework is introduced to initialise and optimise the shape of a dynamic scene for view-dependent rendering, making use of silhouette and stereo data as complementary shape cues. The technique addresses two key problems: (1) robust shape reconstruction; and (2) accurate image correspondence for view dependent rendering in the presence of camera calibration error. We present results against ground truth data in synthetic test cases and for captured sequences of people in a studio. The framework demonstrates a higher resolution in rendering compared to shape from silhouette and multiple view stereo.

    Mohd Azri Mohd Izhar, Marco Volino, Adrian Hilton, Philip Jackson (2020)Tracking Sound Sources for Object-based Spatial Audio in 3D Audio-visual Production, In: Proceedings of the FA2020 Conferencepp. 2051-2058

    In immersive and interactive audio-visual content, there is very significant scope for spatial misalignment between the two main modalities. So, in productions that have both 3D video and spatial audio, the positioning of sound sources relative to the visual display requires careful attention. This may be achieved in the form of object-based audio, moreover allowing the producer to maintain control over individual elements within the mix. Yet each object?s metadata is needed to define its position over time. In the present study, audio-visual studio recordings were made of short scenes representing three genres: drama, sport and music. Foreground video was captured by a light-field camera array, which incorporated a microphone array, alongside more conventional sound recording by spot microphones and a first-order ambisonic room microphone. In the music scenes, a direct feed from the guitar pickup was also recorded. Video data was analysed to form a 3D reconstruction of the scenes, and human figure detection was applied to the 2D frames of the central camera. Visual estimates of the sound source positions were used to provide ground truth. Position metadata were encoded within audio definition model (ADM) format audio files, suitable for standard object-based rendering. The steered response power of the acoustical signals at the microphone array were used, with the phase transform (SRP-PHAT), to determine the dominant source position(s) at any time, and given as input to a Sequential Monte Carlo Probability Hypothesis Density (SMC-PHD) tracker. The output of this was evaluated in relation to the ground truth. Results indicate a hierarchy of accuracy in azimuth, elevation and range, in accordance with human spatial auditory perception. Azimuth errors were within the tolerance bounds reported by studies of the Ventriloquism Effect, giving an initial promising indication that such an approach may open the door to object-based production for live events.

    Matthew Shere, Hansung Kim, Adrian Hilton (2020)3D MULTI PERSON TRACKING WITH DUAL 360°CAMERAS, In: IEEE International Conference on Image Processing

    Person tracking is an often studied facet of computer vision, with applications in security, automated driving and entertainment. However, despite the advantages they offer, few current solutions work for 360° cameras, due to projection distortion. This paper presents a simple yet robust method for 3D tracking of multiple people in a scene from a pair of 360° cameras. By using 2D pose information, rather than potentially unreliable 3D position or repeated colour information, we create a tracker that is both appearance independent as well as capable of operating at narrow baseline. Our results demonstrate state of the art performance on 360° scenes, as well as the capability to handle vertical axis rotation.

    A Neophytou, J-Y Guillemaut, A Hilton (2015)A dense surface motion capture system for accurate acquisition of cloth deformation, In: CVMP 2015: PROCEEDINGS OF THE 12TH EUROPEAN CONFERENCE ON VISUAL MEDIA PRODUCTION
    J Kilner, J-Y Guillemaut, A Hilton (2010)3D action matching with key-pose detection, In: 2009 IEEE 12th International Conference on Computer Vision Workshops, ICCV Workshops 2009pp. 1-8

    This paper addresses the problem of human action matching in outdoor sports broadcast environments, by analysing 3D data from a recorded human activity and retrieving the most appropriate proxy action from a motion capture library. Typically pose recognition is carried out using images from a single camera, however this approach is sensitive to occlusions and restricted fields of view, both of which are common in the outdoor sports environment. This paper presents a novel technique for the automatic matching of human activities which operates on the 3D data available in a multi-camera broadcast environment. Shape is retrieved using multi-camera techniques to generate a 3D representation of the scene. Use of 3D data renders the system camera-pose-invariant and allows it to work while cameras are moving and zooming. By comparing the reconstructions to an appropriate 3D library, action matching can be achieved in the presence of significant calibration and matting errors which cause traditional pose detection schemes to fail. An appropriate feature descriptor and distance metric are presented as well as a technique to use these features for key-pose detection and action matching. The technique is then applied to real footage captured at an outdoor sporting event. ©2009 IEEE.

    HE Imre, J-Y Guillemaut, ADM Hilton (2012)Moving Camera Registration for Multiple Camera Setups in Dynamic Scenes, In: Proceedings of the 21st British Machine Vision Conference

    Many practical applications require an accurate knowledge of the extrinsic calibration (____ie, pose) of a moving camera. The existing SLAM and structure-from-motion solutions are not robust to scenes with large dynamic objects, and do not fully utilize the available information in the presence of static cameras, a common practical scenario. In this paper, we propose an algorithm that addresses both of these issues for a hybrid static-moving camera setup. The algorithm uses the static cameras to build a sparse 3D model of the scene, with respect to which the pose of the moving camera is estimated at each time instant. The performance of the algorithm is studied through extensive experiments that cover a wide range of applications, and is shown to be satisfactory.

    A Hilton, J Illingworth, T Windeatt (1995)Statistics of Surface Curvature Estimates, In: Pattern Recognition288
    J Blat, A Evans, H Kim, H Imre, L Polok, V Ila, N Nikolaidis, P Zamcik, A Tefas, P Smrz, A Hilton, I Pitas (2015)Big Data Analysis for Media Production, In: Proceedings of the IEEE104(11)pp. 2085-2113 IEEE

    A typical high-end film production generates several terabytes of data per day, either as footage from multiple cameras or as background information regarding the set (laser scans, spherical captures, etc). This paper presents solutions to improve the integration, and the understanding of the quality, of the multiple data sources, which are used both to support creative decisions on-set (or near it) and enhance the postproduction process. The main contributions covered in this paper are: a public multisource production dataset made available for research purposes, monitoring and quality assurance of multicamera set-ups, multisource registration, anthropocentric visual analysis for semantic content annotation, acceleration of 3D reconstruction, and integrated 2D-3D web visualization tools. Furthermore, this paper presents a toolset for analysis and visualisation of multi-modal media production datasets which enables onset data quality verification and management, thus significantly reducing the risk and time required in production. Some of the basic techniques used for acceleration, clustering and visualization could be applied to much larger classes of big data problems.

    Graham Miller, Adrian Hilton (2006)Exact View-dependent Visual-hulls, In: ICPRpp. 107-111
    Charles Malleson, Marco Volino, Andrew Gilbert, Matthew Trumble, John Collomosse, Adrian Hilton (2017)Real-time Full-Body Motion Capture from Video and IMUs, In: 3DV 2017 Proceedings CPS

    A real-time full-body motion capture system is presented which uses input from a sparse set of inertial measurement units (IMUs) along with images from two or more standard video cameras and requires no optical markers or specialized infra-red cameras. A real-time optimization-based framework is proposed which incorporates constraints from the IMUs, cameras and a prior pose model. The combination of video and IMU data allows the full 6-DOF motion to be recovered including axial rotation of limbs and drift-free global position. The approach was tested using both indoor and outdoor captured data. The results demonstrate the effectiveness of the approach for tracking a wide range of human motion in real time in unconstrained indoor/outdoor scenes.

    We propose a framework for 2D/3D multi-modal data registration and evaluate 3D feature descriptors for registration of 3D datasets from different sources. 3D datasets of outdoor environments can be acquired using a variety of active and passive sensor technologies including laser scanning and video cameras. Registration of these datasets into a common coordinate frame is required for subsequent modelling and visualisation. 2D images are converted into 3D structure by stereo or multi-view reconstruction techniques and registered to a unified 3D domain with other datasets in a 3D world. Multi-modal datasets have different density, noise, and types of errors in geometry. This paper provides a performance benchmark for existing 3D feature descriptors across multi-modal datasets. Performance is evaluated for the registration of datasets obtained from high-resolution laser scanning with reconstructions obtained from images and video. This analysis highlights the limitations of existing 3D feature detectors and descriptors which need to be addressed for robust multi-modal data registration. We analyse and discuss the performance of existing methods in registering various types of datasets then identify future directions required to achieve robust multi-modal 3D data registration.

    About 1.7 million new cases of breast cancer were estimated by the World Health Organization (WHO) in 2012, accounting for 23 percent of all female cancers. In the UK 33 percent of women aged 50 and above were diagnosed in the same year, thus positioning the UK as the 6th highest in breast cancer amongst the European countries. The national Screening programme in the UK has been focused on the procedure of early detection and to improve prognosis by timely intervention to extend the life span of patients. To this end, the National Health Service Breast Screening Programme (NHSBSP) employs 2-D planar mammography because it is considered to be the gold standard technique for early breast cancer detection worldwide. Breast tomosynthesis has shown great promise as an alternative method for removing the intrinsic overlying clutter seen in conventional 2D imaging. However, preliminary work in breast CT has provided a number of compelling aspects that motivates the work featured in this thesis. These advantages include removal of the need to mechanically compress the breast which is a source of screening non-attendances, and that it provides unique cross sectional images that removes almost all the overlying clutter seen in 2D. This renders lesions more visible and hence aids in early detection of malignancy. However work in Breast CT to date has been focused on using scaled down versions of standard clinical CT systems. By contrast, this thesis proposes using a photon counting approach. The work of this thesis focuses on investigating photoncounting detector technology and comparing it to conventional CT in terms of contrast visualization. Results presented from simulation work developed in this thesis has demonstrated the ability of photoncounting detector technology to utilize data in polychromatic beam where contrast are seen to decrease with increasing photon energy and compared to the conventional CT approach which is the standard clinical CT system.

    J Starck, A Hilton (2003)Towards a 3D Virtual Studio for Human Apperance Capture, In: IMA International Conference on Vision, Video and Graphics, Bath
    C Budd, P Huang, M Klaudiny, A Hilton (2013)Global non-rigid alignment of surface sequences, In: International Journal of Computer Vision102(1-3)pp. 256-270

    This paper presents a general approach based on the shape similarity tree for non-sequential alignment across databases of multiple unstructured mesh sequences from non-rigid surface capture. The optimal shape similarity tree for non-rigid alignment is defined as the minimum spanning tree in shape similarity space. Non-sequential alignment based on the shape similarity tree minimises the total non-rigid deformation required to register all frames in a database into a consistent mesh structure with surfaces in correspondence. This allows alignment across multiple sequences of different motions, reduces drift in sequential alignment and is robust to rapid non-rigid motion. Evaluation is performed on three benchmark databases of 3D mesh sequences with a variety of complex human and cloth motion. Comparison with sequential alignment demonstrates reduced errors due to drift and improved robustness to large non-rigid deformation, together with global alignment across multiple sequences which is not possible with previous sequential approaches. © 2012 The Author(s).

    P Huang, C Budd, A Hilton (2011)Global temporal registration of multiple non-rigid surface sequences, In: Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognitionpp. 3473-3480

    In this paper we consider the problem of aligning multiple non-rigid surface mesh sequences into a single temporally consistent representation of the shape and motion. A global alignment graph structure is introduced which uses shape similarity to identify frames for inter-sequence registration. Graph optimisation is performed to minimise the total non-rigid deformation required to register the input sequences into a common structure. The resulting global alignment ensures that all input sequences are resampled with a common mesh structure which preserves the shape and temporal correspondence. Results demonstrate temporally consistent representation of several public databases of mesh sequences for multiple people performing a variety of motions with loose clothing and hair.

    JJ Kilner, J-Y Guillemaut, A Hilton (2009)3D Action Matching with Key-Pose Detection, In: IEEE 12th International Conference on Computer Vision Workshops (ICCV Workshops)pp. 1-8

    This paper addresses the problem of human action matching in outdoor sports broadcast environments, by analysing 3D data from a recorded human activity and retrieving the most appropriate proxy action from a motion capture library. Typically pose recognition is carried out using images from a single camera, however this approach is sensitive to occlusions and restricted fields of view, both of which are common in the outdoor sports environment. This paper presents a novel technique for the automatic matching of human activities which operates on the 3D data available in a multi-camera broadcast environment. Shape is retrieved using multi-camera techniques to generate a 3D representation of the scene. Use of 3D data renders the system camera-pose-invariant and allows it to work while cameras are moving and zooming. By comparing the reconstructions to an appropriate 3D library, action matching can be achieved in the presence of significant calibration and matting errors which cause traditional pose detection schemes to fail. An appropriate feature descriptor and distance metric are presented as well as a technique to use these features for key-pose detection and action matching. The technique is then applied to real footage captured at an outdoor sporting event

    P Csakany, A Hilton (2006)Relighting of Facial Images, In: IEEE Int.Conf. on Face and Gesture Recognitionpp. 55-60
    J Starck, A Hilton (2006)Free-viewpoint Video for Interactive Character Animation, In: COE Conference, Japan
    D Casas, M Tejera, J-Y Guillemaut, A Hilton (2012)Parametric animation of performance-captured mesh sequences, In: COMPUTER ANIMATION AND VIRTUAL WORLDS23(2)pp. 101-111 WILEY-BLACKWELL
    A Hilton, P Fua, R Ronfard (2006)Vision-based Understanding of a Persons Shape, Appearance, Movement and Behaviour, In: Computer Vision and Image Understanding - Special Issue on Modelling People1042-3
    J Kilner, J Starck, Jean-Yves Guillemaut, A Hilton (2009)Objective Quality Assessment in Free-viewpoint Video Production, In: Signal Processing: Image Communication24(1-2)pp. 3-16 Elsevier
    Charles Malleson, John Collomosse, Adrian Hilton (2019)Real-Time Multi-person Motion Capture from Multi-view Video and IMUs., In: International Journal of Computer Vision Springer

    A real-time motion capture system is presented which uses input from multiple standard video cameras and inertial measurement units (IMUs). The system is able to track multiple people simultaneously and requires no optical markers, specialized infra-red cameras or foreground/background segmentation, making it applicable to general indoor and outdoor scenarios with dynamic backgrounds and lighting. To overcome limitations of prior video or IMU-only approaches, we propose to use flexible combinations of multiple-view, calibrated video and IMU input along with a pose prior in an online optimization-based framework, which allows the full 6-DoF motion to be recovered including axial rotation of limbs and drift-free global position. A method for sorting and assigning raw input 2D keypoint detections into corresponding subjects is presented which facilitates multi-person tracking and rejection of any bystanders in the scene. The approach is evaluated on data from several indoor and outdoor capture environments with one or more subjects and the trade-off between input sparsity and tracking performance is discussed. State-of-the-art pose estimation performance is obtained on the Total Capture (mutli-view video and IMU) and Human 3.6M (multi-view video) datasets. Finally, a live demonstrator for the approach is presented showing real-time capture, solving and character animation using a light-weight, commodity hardware setup.

    P Huang, A Hilton, J Starck (2008)Automatic 3D Video Summarization: Key Frame Extraction from Self-Similarity, In: Proceedings of 3DPVT'08 - the Fourth International Symposium on 3D Data Processing, Visualization and Transmissionpp. 1-8

    In this paper we present an automatic key frame selection method to summarise 3D video sequences. Key-frame selection is based on optimisation for the set of frames which give the best representation of the sequence according to a rate-distortion trade-off. Distortion of the summarization from the original sequence is based on measurement of self-similarity using volume histograms. The method evaluates the globally optimal set of key-frames to represent the entire sequence without requiring pre-segmentation of the sequence into shots or temporal correspondence. Results demonstrate that for 3D video sequences of people wearing a variety of clothing the summarization automatically selects a set of key-frames which represent the dynamics. Comparative evaluation of rate-distortion characteristics with previous 3D video summarization demonstrates improved performance.

    D Casas, Marco Volino, JP Collomosse, A Hilton (2014)4D Video Textures for Interactive Character Appearance, In: Computer Graphics Forum: the international journal of the Eurographics Association

    4D Video Textures (4DVT) introduce a novel representation for rendering video-realistic interactive character animation from a database of 4D actor performance captured in a multiple camera studio. 4D performance capture reconstructs dynamic shape and appearance over time but is limited to free-viewpoint video replay of the same motion. Interactive animation from 4D performance capture has so far been limited to surface shape only. 4DVT is the final piece in the puzzle enabling video-realistic interactive animation through two contributions: a layered view-dependent texture map representation which supports efficient storage, transmission and rendering from multiple view video capture; and a rendering approach that combines multiple 4DVT sequences in a parametric motion space, maintaining video quality rendering of dynamic surface appearance whilst allowing high-level interactive control of character motion and viewpoint. 4DVT is demonstrated for multiple characters and evaluated both quantitatively and through a user-study which confirms that the visual quality of captured video is maintained. The 4DVT representation achieves >90% reduction in size and halves the rendering cost.

    J Illingworth, A Hilton (1998)Looking to Build a Model World: Automatic Construction of Static Object Models using Computer Vision, In: IEE Journal Electronics and Communications Engineering103
    Charles Malleson, Jean-Yves Guillemaut, Adrian Hilton (2019)3D Reconstruction from RGB-D Data, In: Paul L. Rosin, Yu-Kun Lai, Ling Shao, Yonghuai Liu (eds.), RGB-D Image Analysis and Processingpp. pp 87-115 Springer Nature Switzerland AG

    A key task in computer vision is that of generating virtual 3D models of real-world scenes by reconstructing the shape, appearance and, in the case of dynamic scenes, motion of the scene from visual sensors. Recently, low-cost video plus depth (RGB-D) sensors have become widely available and have been applied to 3D reconstruction of both static and dynamic scenes. RGB-D sensors contain an active depth sensor, which provides a stream of depth maps alongside standard colour video. The low cost and ease of use of RGB-D devices as well as their video rate capture of images along with depth make them well suited to 3D reconstruction. Use of active depth capture overcomes some of the limitations of passive monocular or multiple-view video-based approaches since reliable, metrically accurate estimates of the scene depth at each pixel can be obtained from a single view, even in scenes that lack distinctive texture. There are two key components to 3D reconstruction from RGB-D data: (1) spatial alignment of the surface over time and, (2) fusion of noisy, partial surface measurements into a more complete, consistent 3D model. In the case of static scenes, the sensor is typically moved around the scene and its pose is estimated over time. For dynamic scenes, there may be multiple rigid, articulated, or non-rigidly deforming surfaces to be tracked over time. The fusion component consists of integration of the aligned surface measurements, typically using an intermediate representation, such as the volumetric truncated signed distance field (TSDF). In this chapter, we discuss key recent approaches to 3D reconstruction from depth or RGB-D input, with an emphasis on real-time reconstruction of static scenes.

    M Hamouz, JR Tena, J Kittler, A Hilton, J Illingworth (2006)Algorithms for 3D-Assisted Face Recognition, In: IEEE 14th Signal Processing and Communications Applications Conference ( SIU06)
    Samuel Fowler, Hansung Kim, Adrian Hilton (2017)Towards Complete Scene Reconstruction from Single-View Depth and Human Motion, In: Proceedings of the 28th British Machine Vision Conference (BMVC 2017)

    Complete scene reconstruction from single view RGBD is a challenging task, requiring estimation of scene regions occluded from the captured depth surface. We propose that scene-centric analysis of human motion within an indoor scene can reveal fully occluded objects and provide functional cues to enhance scene understanding tasks. Captured skeletal joint positions of humans, utilised as naturally exploring active sensors, are projected into a human-scene motion representation. Inherent body occupancy is leveraged to carve a volumetric scene occupancy map initialised from captured depth, revealing a more complete voxel representation of the scene. To obtain a structured box model representation of the scene, we introduce unique terms to an object detection optimisation that overcome depth occlusions whilst deriving from the same depth data. The method is evaluated on challenging indoor scenes with multiple occluding objects such as tables and chairs. Evaluation shows that human-centric scene analysis can be applied to effectively enhance state-of-the-art scene understanding approaches, resulting in a more complete representation than single view depth alone.

    H Kim, A Hilton (2009)GRAPH-BASED FOREGROUND EXTRACTION IN EXTENDED COLOR SPACE, In: 2009 16TH IEEE INTERNATIONAL CONFERENCE ON IMAGE PROCESSING, VOLS 1-6pp. 3221-3224
    A Hilton (1999)Towards Model-based Capture of a Persons Shape, Appearance and Motion, In: IEEE International Workshop on Modelling People

    This paper introduces a model-based approach to capturing a persons shape, appearance and movement. A 3D animated model of a clothed persons whole-body shape and appearance is automatically constructed from a set of orthogonal view colour images. The reconstructed model of a person is then used together with the least-squares inverse-kinematics framework of Bregler and Malik (1998) to capture simple 3D movements from a video image sequence

    We propose a multi-view framework for joint object detection and labelling based on pairs of images. The proposed framework extends the single-view Mask R-CNN approach to multiple views without need for additional training. Dedicated components are embedded into the framework to match objects across views by enforcing epipolar constraints, appearance feature similarity and class coherence. The multi-view extension enables the proposed framework to detect objects which would otherwise be mis-detected in a classical Mask R-CNN approach, and achieves coherent object labelling across views. By avoiding the need for additional training, the approach effectively overcomes the current shortage of multi-view datasets. The proposed framework achieves high quality results on a range of complex scenes, being able to output class, bounding box, mask and an additional label enforcing coherence across views. In the evaluation, we show qualitative and quantitative results on several challenging outd oor multi-view datasets and perform a comprehensive comparison to verify the advantages of the proposed method

    Armin Mustafa, Adrian Hilton (2019)Semantically Coherent 4D Scene Flow of Dynamic Scenes, In: International Journal of Computer Vision Springer Verlag

    Simultaneous semantically coherent object-based long-term 4D scene flow estimation, co-segmentation and reconstruction is proposed exploiting the coherence in semantic class labels both spatially, between views at a single time instant, and temporally, between widely spaced time instants of dynamic objects with similar shape and appearance. In this paper we propose a framework for spatially and temporally coherent semantic 4D scene flow of general dynamic scenes from multiple view videos captured with a network of static or moving cameras. Semantic coherence results in improved 4D scene flow estimation, segmentation and reconstruction for complex dynamic scenes. Semantic tracklets are introduced to robustly initialize the scene flow in the joint estimation and enforce temporal coherence in 4D flow, semantic labelling and reconstruction between widely spaced instances of dynamic objects. Tracklets of dynamic objects enable unsupervised learning of long-term flow, appearance and shape priors that are exploited in semantically coherent 4D scene flow estimation, co-segmentation and reconstruction. Comprehensive performance evaluation against state-of-the-art techniques on challenging indoor and outdoor sequences with hand-held moving cameras shows improved accuracy in 4D scene flow, segmentation, temporally coherent semantic labelling, and reconstruction of dynamic scenes.

    JD Edge, A Hilton (2007)Facial Animation with Motion Capture based on Surface Blending, In: International Conference on Computer Graphics Theory and Applications
    IA Ypsilos, A Hilton, A Turkmani, PJB Jackson (2004)Speech Driven Face Synthesis from 3D Video, In: IEEE Symposium on 3D Data Processing, Visualisation and Transmissionpp. 58-65

    We present a framework for speech-driven synthesis of real faces from a corpus of 3D video of a person speaking. Video-rate capture of dynamic 3D face shape and colour appearance provides the basis for a visual speech synthesis model. A displacement map representation combines face shape and colour into a 3D video. This representation is used to efficiently register and integrate shape and colour information captured from multiple views. To allow visual speech synthesis viseme primitives are identified from the corpus using automatic speech recognition. A novel nonrigid alignment algorithm is introduced to estimate dense correspondence between 3D face shape and appearance for different visemes. The registered displacement map representation together with a novel optical flow optimisation using both shape and colour, enables accurate and efficient nonrigid alignment. Face synthesis from speech is performed by concatenation of the corresponding viseme sequence using the nonrigid correspondence to reproduce both 3D face shape and colour appearance. Concatenative synthesis reproduces both viseme timing and co-articulation. Face capture and synthesis has been performed for a database of 51 people. Results demonstrate synthesis of 3D visual speech animation with a quality comparable to the captured video of a person.

    P Huang, A Hilton, J Starck (2009)Human Motion Synthesis from 3D Video, In: Proceedings of IEEE Conference on Computer Vision and Pattern Recognitionpp. 1478-1485

    Multiple view 3D video reconstruction of actor performance captures a level-of-detail for body and clothing movement which is time-consuming to produce using existing animation tools. In this paper we present a framework for concatenative synthesis from multiple 3D video sequences according to user constraints on movement, position and timing. Multiple 3D video sequences of an actor performing different movements are automatically constructed into a surface motion graph which represents the possible transitions with similar shape and motion between sequences without unnatural movement artifacts. Shape similarity over an adaptive temporal window is used to identify transitions between 3D video sequences. Novel 3D video sequences are synthesized by finding the optimal path in the surface motion graph between user specified key-frames for control of movement, location and timing. The optimal path which satisfies the user constraints whilst minimizing the total transition cost between 3D video sequences is found using integer linear programming. Results demonstrate that this framework allows flexible production of novel 3D video sequences which preserve the detailed dynamics of the captured movement for an actress with loose clothing and long hair without visible artifacts.

    Matthew Trumble, Andrew Gilbert, Adrian Hilton, John Collomosse (2016)Learning Markerless Human Pose Estimation from Multiple Viewpoint Video, In: Computer Vision – ECCV 2016 Workshops. Lecture Notes in Computer Science9915pp. 871-878

    We present a novel human performance capture technique capable of robustly estimating the pose (articulated joint positions) of a performer observed passively via multiple view-point video (MVV). An affine invariant pose descriptor is learned using a convolutional neural network (CNN) trained over volumetric data extracted from a MVV dataset of diverse human pose and appearance. A manifold embedding is learned via Gaussian Processes for the CNN descriptor and articulated pose spaces enabling regression and so estimation of human pose from MVV input. The learned descriptor and manifold are shown to generalise over a wide range of human poses, providing an efficient performance capture solution that requires no fiducials or other markers to be worn. The system is evaluated against ground truth joint configuration data from a commercial marker-based pose estimation system

    J Mitchelson, A Hilton (2003)Hierarchical Tracking of Human Motion for Animation, In: Model-based Imaging, Rendering, image Analysis and Graphical Special Effects, Paris
    A Doshi, A Hilton, J Starck (2008)An Empirical Study of Non-rigid Surface Feature Matching, In: European Conference on Visual Media Production
    A Hilton, J Illingworth, Y Li, J Mitchelson (2017)Real-Time Human Motion Estimation for Studio Production, In: BMVA Workshop on Understanding Human Behaviour
    A Mustafa, H Kim, H Imre, A Hilton (2015)Segmentation based features for wide-baseline multi-view reconstruction, In: International Conference on 3D Vision (3DV)pp. 282-290

    A common problem in wide-baseline stereo is the sparse and non-uniform distribution of correspondences when using conventional detectors such as SIFT, SURF, FAST and MSER. In this paper we introduce a novel segmentation based feature detector SFD that produces an increased number of ‘good’ features for accurate wide-baseline reconstruction. Each image is segmented into regions by over-segmentation and feature points are detected at the intersection of the boundaries for three or more regions. Segmentation-based feature detection locates features at local maxima giving a relatively large number of feature points which are consistently detected across wide-baseline views and accurately localised. A comprehensive comparative performance evaluation with previous feature detection approaches demonstrates that: SFD produces a large number of features with increased scene coverage; detected features are consistent across wide-baseline views for images of a variety of indoor and outdoor scenes; and the number of wide-baseline matches is increased by an order of magnitude compared to alternative detector-descriptor combinations. Sparse scene reconstruction from multiple wide-baseline stereo views using the SFD feature detector demonstrates at least a factor six increase in the number of reconstructed points with reduced error distribution compared to SIFT when evaluated against ground-truth and similar computational cost to SURF/FAST.

    Charles Malleson, Jean-Charles Bazin, Oliver Wang, Derek Bradley, Thabo Beeler, Adrian Hilton, Alexander Sorkine-Hornung (2016)FaceDirector: Continuous Control of Facial Performance in Video, In: 2015 IEEE International Conference on Computer Vision (ICCV)pp. 3979-3987 Institute of Electrical and Electronics Engineers (IEEE)

    We present a method to continuously blend between multiple facial performances of an actor, which can contain different facial expressions or emotional states. As an example, given sad and angry video takes of a scene, our method empowers the movie director to specify arbitrary weighted combinations and smooth transitions between the two takes in post-production. Our contributions include (1) a robust nonlinear audio-visual synchronization technique that exploits complementary properties of audio and visual cues to automatically determine robust, dense spatiotemporal correspondences between takes, and (2) a seamless facial blending approach that provides the director full control to interpolate timing, facial expression, and local appearance, in order to generate novel performances after filming. In contrast to most previous works, our approach operates entirely in image space, avoiding the need of 3D facial reconstruction. We demonstrate that our method can synthesize visually believable performances with applications in emotion transition, performance correction, and timing control.

    D Cosker, E Krumhuber, A Hilton (2010)Perception of Linear and Nonlinear Motion Properties using a FACS Validated 3D Facial Model, In: In Proc. of ACM Symposium on Applied Perception in Graphics and Visualisation (APGV)pp. 101-108

    In this paper we present the first Facial Action Coding System (FACS) valid model to be based on dynamic 3D scans of human faces for use in graphics and psychological research. The model consists of FACS Action Unit (AU) based parameters and has been independently validated by FACS experts. Using this model, we explore the perceptual differences between linear facial motions – represented by a linear blend shape approach – and real facial motions that have been synthesized through the 3D facial model. Through numerical measures and visualizations, we show that this latter type of motion is geometrically nonlinear in terms of its vertices. In experiments, we explore the perceptual benefits of nonlinear motion for different AUs. Our results are insightful for designers of animation systems both in the entertainment industry and in scientific research. They reveal a significant overall benefit to using captured nonlinear geometric vertex motion over linear blend shape motion. However, our findings suggest that not all motions need to be animated nonlinearly. The advantage may depend on the type of facial action being produced and the phase of the movement.

    LM Tanco, A Hilton (2000)Realistic synthesis of novel human movements from a database of motion capture examples, In: Workshop on Workshop on Human Motion, 2000.pp. 137-142
    A Saminathan, AJ Stoddart, A Hilton, J Illingworth (1997)Progress in arbitrary topology deformable surfaces, In: BMVC
    T Wang, P McLauchlan, P Palmer, A Hilton (2001)Calibration for an Integrated Measurement System of Camera and Laser and its Application, In: 5th World Multiconference on Systemics, Cybernetics and Informatics (Awarded Best Paper), Orlando, Florida, USA
    J Starck, A Hilton (2007)Surface Capture for Performance-Based Animation, In: IEEE Computer Graphics and Applications27(3)pp. 21-31 IEEE
    A Hilton, AJ Stoddart, J Illingworth, T Windeatt (1996)Implicit Surface based Geometric Fusion, In: Leeds 16th Annual Statistics Workshoppp. 1-8
    Hansung Kim, T de Campos, Adrian Hilton (2016)Room Layout Estimation with Object and Material Attributes Information using a Spherical Camera, In: Fourth International Conference on 3D Vision (3DV)

    In this paper we propose a pipeline for estimating 3D room layout with object and material attribute prediction using a spherical stereo image pair. We assume that the room and objects can be represented as cuboids aligned to the main axes of the room coordinate (Manhattan world). A spherical stereo alignment algorithm is proposed to align two spherical images to the global world coordinate sys- tem. Depth information of the scene is estimated by stereo matching between images. Cubic projection images of the spherical RGB and estimated depth are used for object and material attribute detection. A single Convolutional Neu- ral Network is designed to assign object and attribute la- bels to geometrical elements built from the spherical image. Finally simplified room layout is reconstructed by cuboid fitting. The reconstructed cuboid-based model shows the structure of the scene with object information and material attributes.

    N Nadtoka, A Hilton, J Tena, J Edge, PJB Jackson (2006)Representing Dynamics of Facial Expression, In: IET European Conference on Visual Media Productionpp. 183-183

    Motion capture (mocap) is widely used in a large number of industrial applications. Our work offers a new way of representing the mocap facial dynamics in a high resolution 3D morphable model expression space. A data-driven approach to modelling of facial dynamics is presented. We propose a way to combine high quality static face scans with dynamic 3D mocap data which has lower spatial resolution in order to study the dynamics of facial expressions.

    A Hilton, J Starck (2004)Multiple View Reconstruction of People, In: IEEE Conference on 3D Data Processing, Visualisation and Transmissionpp. 357-364
    A Hilton, J Starck (2007)Animation of People from Surface Motion Capture, In: IEEE Computer Graphics and Applications3(27)pp. 21-31

    Digital content production traditionally requires highly skilled artists and animators to first manually craft shape and appearance models and then instill the models with a believable performance. Motion capture technology is now increasingly used to record the articulated motion of a real human performance to increase the visual realism in animation. Motion capture is limited to recording only the skeletal motion of the human body and requires the use of specialist suits and markers to track articulated motion. In this paper we present surface capture, a fully automated system to capture shape and appearance as well as motion from multiple video cameras as a basis to create highly realistic animated content from an actor’s performance in full wardrobe. We address wide-baseline scene reconstruction to provide 360 degree appearance from just 8 camera views and introduce an efficient scene representation for level of detail control in streaming and rendering. Finally we demonstrate interactive animation control in a computer games scenario using a captured library of human animation, achieving a frame rate of 300fps on consumer level graphics hardware.

    IA Ypsilos, A Hilton, S Rowe (2004)Video-rate Capture of Dynamic Face Shape and Appearance, In: IEEE Face and Gesture Recognition

    This paper presents a system for simultaneous capture of video sequences of face shape and colour appearance. Shape capture uses a projected infra-red structured light pattern together with stereo reconstruction to simultaneously acquire full resolution shape and colour image sequences at video rate. Displacement mapping techniques are introduced to represent dynamic face surface shape as a displacement video. This unifies the representation of face shape and colour. The displacement video representation enables efficient registration, integration and spatiotemporal analysis of captured face data. Results demonstrate that the system achieves video-rate (25Hz) acquisition of dynamic 3D colour faces at PAL resolution with an rms accuracy of 0.2mm and a visual quality comparable to the captured video.

    MATTHEW JAMES SHERE, HANSUNG KIM, ADRIAN DOUGLAS MARK HILTON (2020)Temporally Consistent 3D Human Pose Estimation Using Dual 360° Cameras

    This paper presents a 3D human pose estimation system that uses a stereo pair of 360° sensors to capture the complete scene from a single location. The approach combines the advantages of omnidirectional capture, the accuracy of multiple view 3D pose estimation and the portability of monocular acquisition. Joint monocular belief maps for joint locations are estimated from 360° images and are used to fit a 3D skeleton to each frame. Temporal data association and smoothing is performed to produce accurate 3D pose estimates throughout the sequence. We evaluate our system on the Panoptic Studio dataset, as well as real 360° video for tracking multiple people, demonstrating an average Mean Per Joint Position Error of 12.47cm with 30cm baseline cameras. We also demonstrate improved capabilities over perspective and 360° multi-view systems when presented with limited camera views of the subject.

    A Ahmed, A Hilton, F Mokhtarian (2004)Enriching Animation Databases, In: Eurographics Short Paper
    Armin Mustafa, Christopher Russell, Adrian Hilton (2019)U4D: Unsupervised 4D Dynamic Scene Understanding, In: Proceedings 2019 IEEE International Conference on Computer Vision IEEE

    We introduce the first approach to solve the challenging problem of unsupervised 4D visual scene understanding for complex dynamic scenes with multiple interacting people from multi-view video. Our approach simultaneously estimates a detailed model that includes a per-pixel semantically and temporally coherent reconstruction, together with instance-level segmentation exploiting photo-consistency, semantic and motion information. We further leverage recent advances in 3D pose estimation to constrain the joint semantic instance segmentation and 4D temporally coherent reconstruction. This enables per person semantic instance segmentation of multiple interacting people in complex dynamic scenes. Extensive evaluation of the joint visual scene understanding framework against state-of-the-art methods on challenging indoor and outdoor sequences demonstrates a significant (≈ 40%) improvement in semantic segmentation, reconstruction and scene flow accuracy.

    Hansung Kim, Luca Remaggi, Philip Jackson, Adrian Hilton (2019)Immersive Spatial Audio Reproduction for VR/AR Using Room Acoustic Modelling from 360° Images, In: Proceedings IEEE VR2019 IEEE

    Recent progresses in Virtual Reality (VR) and Augmented Reality (AR) allow us to experience various VR/AR applications in our daily life. In order to maximise the immersiveness of user in VR/AR environments, a plausible spatial audio reproduction synchronised with visual information is essential. In this paper, we propose a simple and efficient system to estimate room acoustic for plausible reproducton of spatial audio using 360° cameras for VR/AR applications. A pair of 360° images is used for room geometry and acoustic property estimation. A simplified 3D geometric model of the scene is estimated by depth estimation from captured images and semantic labelling using a convolutional neural network (CNN). The real environment acoustics are characterised by frequency-dependent acoustic predictions of the scene. Spatially synchronised audio is reproduced based on the estimated geometric and acoustic properties in the scene. The reconstructed scenes are rendered with synthesised spatial audio as VR/AR content. The results of estimated room geometry and simulated spatial audio are evaluated against the actual measurements and audio calculated from ground-truth Room Impulse Responses (RIRs) recorded in the rooms.

    J Imber, J-Y Guillemaut, A Hilton (2014)Intrinsic textures for relightable free-viewpoint video, In: Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)8690 L(PART 2)pp. 392-407

    This paper presents an approach to estimate the intrinsic texture properties (albedo, shading, normal) of scenes from multiple view acquisition under unknown illumination conditions. We introduce the concept of intrinsic textures, which are pixel-resolution surface textures representing the intrinsic appearance parameters of a scene. Unlike previous video relighting methods, the approach does not assume regions of uniform albedo, which makes it applicable to richly textured scenes. We show that intrinsic image methods can be used to refine an initial, low-frequency shading estimate based on a global lighting reconstruction from an original texture and coarse scene geometry in order to resolve the inherent global ambiguity in shading. The method is applied to relighting of free-viewpoint rendering from multiple view video capture. This demonstrates relighting with reproduction of fine surface detail. Quantitative evaluation on synthetic models with textured appearance shows accurate estimation of intrinsic surface reflectance properties. © 2014 Springer International Publishing.

    J-Y Guillemaut, A Hilton (2012)Space-Time Joint Multi-Layer Segmentation and Depth Estimation, In: SECOND JOINT 3DIM/3DPVT CONFERENCE: 3D IMAGING, MODELING, PROCESSING, VISUALIZATION & TRANSMISSION (3DIMPVT 2012)pp. 440-447
    A Hilton (2003)Computer Vision for Human Modelling and Analysis, In: Journal of Machine Vision Applications144
    A Hilton (1992)Algorithms for Estimating Turbulent Flow Parameters from In-Cylinder Laser Doppler Anemometer Data Doctor of Philosophy (D.Phil.) Thesis, University of Sussex,UK
    L Molina, A Hilton (2001)Learning models for sythesis of human motion, In: BMVA Workshop on Probabalistic Methods in Computer Vision
    M Sarim, JY Guillemaut, H Kim, A Hilton (2009)Wide-baseline Image Matting, In: European Conference on Visual Media Production(CVMP)
    Matthew Trumble, Andrew Gilbert, Charles Malleson, Adrian Hilton, John Collomosse (2017)Total Capture: 3D Human Pose Estimation Fusing Video and Inertial Sensors, In: Proceedings of 28th British Machine Vision Conferencepp. 1-13

    We present an algorithm for fusing multi-viewpoint video (MVV) with inertial measurement unit (IMU) sensor data to accurately estimate 3D human pose. A 3-D convolutional neural network is used to learn a pose embedding from volumetric probabilistic visual hull data (PVH) derived from the MVV frames. We incorporate this model within a dual stream network integrating pose embeddings derived from MVV and a forward kinematic solve of the IMU data. A temporal model (LSTM) is incorporated within both streams prior to their fusion. Hybrid pose inference using these two complementary data sources is shown to resolve ambiguities within each sensor modality, yielding improved accuracy over prior methods. A further contribution of this work is a new hybrid MVV dataset (TotalCapture) comprising video, IMU and a skeletal joint ground truth derived from a commercial motion capture system. The dataset is available online at http://cvssp.org/data/totalcapture/.

    J Woodcock, C Pike, F Melchior, Philip Coleman, A Franck, Adrian Hilton (2016)Presenting the S3A Object-Based Audio Drama dataset, In: AES E-library Audio Engineering Society

    This engineering brief reports on the production of 3 object-based audio drama scenes, commissioned as part of the S3A project. 3D reproduction and an object-based workflow were considered and implemented from the initial script commissioning through to the final mix of the scenes. The scenes are being made available as Broadcast Wave Format files containing all objects as separate tracks and all metadata necessary to render the scenes as an XML chunk in the header conforming to the Audio Definition Model specification (Recommendation ITU-R BS.2076 [1]). It is hoped that these scenes will find use in perceptual experiments and in the testing of 3D audio systems. The scenes are available via the following link: http://dx.doi.org/10.17866/rd.salford.3043921.

    John Collomosse, P Huang, A Hilton, M Tejera (2015)Hybrid Skeletal-Surface Motion Graphs for Character Animation from 4D Performance Capture, In: ACM Transactions on Graphics34(2)

    We present a novel hybrid representation for character animation from 4D Performance Capture (4DPC) data which combines skeletal control with surface motion graphs. 4DPC data are temporally aligned 3D mesh sequence reconstructions of the dynamic surface shape and associated appearance from multiple view video. The hybrid representation supports the production of novel surface sequences which satisfy constraints from user specified key-frames or a target skeletal motion. Motion graph path optimisation concatenates fragments of 4DPC data to satisfy the constraints whilst maintaining plausible surface motion at transitions between sequences. Spacetime editing of the mesh sequence using a learnt part-based Laplacian surface deformation model is performed to match the target skeletal motion and transition between sequences. The approach is quantitatively evaluated for three 4DPC datasets with a variety of clothing styles. Results for key-frame animation demonstrate production of novel sequences which satisfy constraints on timing and position of less than 1% of the sequence duration and path length. Evaluation of motion capture driven animation over a corpus of 130 sequences shows that the synthesised motion accurately matches the target skeletal motion. The combination of skeletal control with the surface motion graph extends the range and style of motion which can be produced whilst maintaining the natural dynamics of shape and appearance from the captured performance.

    P Csakany, F Vajda, A Hilton (2007)Recovering Refined Surface Normals for Relighting Clothing in Dynamic Scenes, In: IET European Conference on Visual Media Production

    In this paper we present a method to relight captured 3D video sequences of non-rigid, dynamic scenes, such as clothing of real actors, reconstructed from multiple view video. A view-dependent approach is introduced to refine an initial coarse surface reconstruction using shape-from-shading to estimate detailed surface normals. The prior surface approximation is used to constrain the simultaneous estimation of surface normals and scene illumination, under the assumption of Lambertian surface reflectance. This approach enables detailed surface normals of a moving non-rigid object to be estimated from a single image frame. Refined normal estimates from multiple views are integrated into a single surface normal map. This approach allows highly non-rigid surfaces, such as creases in clothing, to be relit whilst preserving the detailed dynamics observed in video.

    P Csakany, A Hilton (2006)Relighting of Facial Video, In: ICPRpp. 203-206

    We present a novel method to relight video sequences given known surface shape and illumination. The method preserves fine visual details. It requires single view video frames, approximate 3D shape and standard studio illumination only, making it applicable in studio production. The technique is demonstrated for relighting video sequences of faces

    JD Edge, A Hilton, PJB Jackson (2009)Model-based synthesis of visual speech movements from 3D video, In: EURASIP Journal on Audio, Speech, and Music Processing2009597267pp. 12-12 Hindawi Publishing Corporation

    In this paper we describe a method for the synthesis of visual speech movements using a hybrid unit selection/model-based approach. Speech lip movements are captured using a 3D stereo face capture system, and split up into phonetic units. A dynamic parameterisation of this data is constructed which maintains the relationship between lip shapes and velocities; within this parameterisation a model of how lips move is built and is used in the animation of visual speech movements from speech audio input. The mapping from audio parameters to lip movements is disambiguated by selecting only the most similar stored phonetic units to the target utterance during synthesis. By combining properties of model-based synthesis (e.g. HMMs, neural nets) with unit selection we improve the quality of our speech synthesis.

    A Stoddart, S Lemke, A Hilton, T Renn (1998)Estimating pose uncertainty for surface registration, In: Image and Vision Computing162pp. 111-120
    J Starck, A Hilton (2005)Spherical Matching for Temporal Correspondence of Non-Rigid Surfaces, In: IEEE Int.Conf.Computer Vision2pp. 1387-1394
    D Casas, M Tejera, Jean-Yves Guillemaut, A Hilton (2012)4D parametric motion graphs for interactive animation, In: I3D '12 Proceedings of the ACM SIGGRAPH Symposium on Interactive 3D Graphics and Gamespp. 103-110 ACM

    A 4D parametric motion graph representation is presented for interactive animation from actor performance capture in a multiple camera studio. The representation is based on a 4D model database of temporally aligned mesh sequence reconstructions for multiple motions. High-level movement controls such as speed and direction are achieved by blending multiple mesh sequences of related motions. A real-time mesh sequence blending approach is introduced which combines the realistic deformation of previous non-linear solutions with efficient online computation. Transitions between different parametric motion spaces are evaluated in real-time based on surface shape and motion similarity. 4D parametric motion graphs allow real-time interactive character animation while preserving the natural dynamics of the captured performance. © 2012 ACM.

    M Sarim, A Hilton, Jean-Yves Guillemaut, Hansung Kim, T Takai (2010)Multiple view wide-baseline trimap propagation for natural video matting, In: Proc. European Conference on Visual Media Production (CVMP 2010)pp. 82-91

    This paper presents a method to estimate alpha mattes for video sequences of the same foreground scene from wide-baseline views given sparse key-frame trimaps in a single view. A statistical inference framework is introduced for spatio-temporal propagation of high-confidence trimap labels between video sequences without a requirement for correspondence or camera calibration and motion estimation. Multiple view trimap propagation integrates appearance information between views and over time to achieve robust labelling in the presence of shadows, changes in appearance with view point and overlap between foreground and background appearance. Results demonstrate that trimaps are sufficiently accurate to allow high-quality video matting using existing single view natural image matting algorithms. Quantitative evaluation against ground-truth demonstrates that the approach achieves accurate matte estimation for camera views separated by up to 180◦ , with the same amount of manual interaction required for conventional single view video matting.

    V Kilic, M Barnard, W Wang, A Hilton, J Kittler (2016)Mean-Shift and Sparse Sampling Based SMC-PHD Filtering for Audio Informed Visual Speaker Tracking, In: IEEE Transactions on Multimedia

    The probability hypothesis density (PHD) filter based on sequential Monte Carlo (SMC) approximation (also known as SMC-PHD filter) has proven to be a promising algorithm for multi-speaker tracking. However, it has a heavy computational cost as surviving, spawned and born particles need to be distributed in each frame to model the state of the speakers and to estimate jointly the variable number of speakers with their states. In particular, the computational cost is mostly caused by the born particles as they need to be propagated over the entire image in every frame to detect the new speaker presence in the view of the visual tracker. In this paper, we propose to use audio data to improve the visual SMC-PHD (VSMC- PHD) filter by using the direction of arrival (DOA) angles of the audio sources to determine when to propagate the born particles and re-allocate the surviving and spawned particles. The tracking accuracy of the AV-SMC-PHD algorithm is further improved by using a modified mean-shift algorithm to search and climb density gradients iteratively to find the peak of the probability distribution, and the extra computational complexity introduced by mean-shift is controlled with a sparse sampling technique. These improved algorithms, named as AVMS-SMCPHD and sparse-AVMS-SMC-PHD respectively, are compared systematically with AV-SMC-PHD and V-SMC-PHD based on the AV16.3, AMI and CLEAR datasets.

    Matthew Trumble, Andrew Gilbert, Adrian Hilton, John Collomosse (2018)Deep Autoencoder for Combined Human Pose Estimation and Body Model Upscaling, In: Proceedings of ECCV 2018: European Conference on Computer Vision Springer Science+Business Media

    We present a method for simultaneously estimating 3D hu- man pose and body shape from a sparse set of wide-baseline camera views. We train a symmetric convolutional autoencoder with a dual loss that enforces learning of a latent representation that encodes skeletal joint positions, and at the same time learns a deep representation of volumetric body shape. We harness the latter to up-scale input volumetric data by a factor of 4X, whilst recovering a 3D estimate of joint positions with equal or greater accuracy than the state of the art. Inference runs in real-time (25 fps) and has the potential for passive human behaviour monitoring where there is a requirement for high fidelity estimation of human body shape and pose.

    Joao Regateiro, Marco Volino, Adrian Hilton (2018)Hybrid Skeleton Driven Surface Registration for Temporally Consistent Volumetric Video, In: Proceedings of 2018 International Conference on 3D Vision (3DV)pp. 514-522 Institute of Electrical and Electronics Engineers (IEEE)

    This paper presents a hybrid skeleton-driven surface registration (HSDSR) approach to generate temporally consistent meshes from multiple view video of human subjects. 2D pose detections from multiple view video are used to estimate 3D skeletal pose on a per-frame basis. The 3D pose is embedded into a 3D surface reconstruction allowing any frame to be reposed into the shape from any other frame in the captured sequence. Skeletal motion transfer is performed by selecting a reference frame from the surface reconstruction data and reposing it to match the pose estimation of other frames in a sequence. This allows an initial coarse alignment to be performed prior to refinement by a patch-based non-rigid mesh deformation. The proposed approach overcomes limitations of previous work by reposing a reference mesh to match the pose of a target mesh reconstruction, providing a closer starting point for further non-rigid mesh deformation. It is shown that the proposed approach is able to achieve comparable results to existing model-based and model-free approaches. Finally, it is demonstrated that this framework provides an intuitive way for artists and animators to edit volumetric video.

    Qingju Liu, Wenwu Wang, Teofilo de Campos, Philip Jackson, Adrian Hilton (2017)Multiple Speaker Tracking in Spatial Audio via PHD Filtering and Depth-Audio Fusion, In: IEEE Transactions on Multimedia20(7)pp. 1767-1780 Institute of Electrical and Electronics Engineers (IEEE)

    In object-based spatial audio system, positions of the audio objects (e.g. speakers/talkers or voices) presented in the sound scene are required as important metadata attributes for object acquisition and reproduction. Binaural microphones are often used as a physical device to mimic human hearing and to monitor and analyse the scene, including localisation and tracking of multiple speakers. The binaural audio tracker, however, is usually prone to the errors caused by room reverberation and background noise. To address this limitation, we present a multimodal tracking method by fusing the binaural audio with depth information (from a depth sensor, e.g., Kinect). More specifically, the PHD filtering framework is first applied to the depth stream, and a novel clutter intensity model is proposed to improve the robustness of the PHD filter when an object is occluded either by other objects or due to the limited field of view of the depth sensor. To compensate mis-detections in the depth stream, a novel gap filling technique is presented to map audio azimuths obtained from the binaural audio tracker to 3D positions, using speaker-dependent spatial constraints learned from the depth stream. With our proposed method, both the errors in the binaural tracker and the mis-detections in the depth tracker can be significantly reduced. Real-room recordings are used to show the improved performance of the proposed method in removing outliers and reducing mis-detections.

    Graham Miller, JR Starck, Adrian Hilton (2006)Projective Surface Refinement for Free-Viewpoint Video, In: IET European Conference on Visual Media Productionpp. 153-162

    This paper introduces a novel method of surface refinement for free-viewpoint video of dynamic scenes. Unlike previous approaches, the method presented here uses both visual hull and silhouette contours to constrain refinement of viewdependent depth maps from wide baseline views. A technique for extracting silhouette contours as rims in 3D from the view-dependent visual hull (VDVH) is presented. A new method for improving correspondence is introduced, where refinement of the VDVH is posed as a global problem in projective ray space. Artefacts of global optimisations are reduced by incorporating rims as constraints. Real time rendering of virtual views in a free-viewpoint video system is achieved using an image+depth representation for each real view. Results illustrate the high quality of rendered views achieved through this refinement technique.

    A Hilton, AJ Stoddart, J Illingworth, T Windeatt (1996)Building 3D Graphical Models of Complex Objects, In: Eurographics UK Conference
    J Starck, A Hilton (2003)Model-based Multiple View Reconstruction of People, In: IEEE International Conference on Computer Visionpp. 915-922
    W Sun, A Hilton, R Smith, J Illingworth (2001)Layered Animation of Captured Data, In: Visual Computer: International Journal of Computer Graphics17(8)pp. 457-474 Springer
    Armin Mustafa, Adrian Hilton (2017)Semantically Coherent Co-segmentation and Reconstruction of Dynamic Scenes, In: CVPR 2017 Proceedingspp. 5583-5592 IEEE

    In this paper we propose a framework for spatially and temporally coherent semantic co-segmentation and reconstruction of complex dynamic scenes from multiple static or moving cameras. Semantic co-segmentation exploits the coherence in semantic class labels both spatially, between views at a single time instant, and temporally, between widely spaced time instants of dynamic objects with similar shape and appearance. We demonstrate that semantic coherence results in improved segmentation and reconstruction for complex scenes. A joint formulation is proposed for semantically coherent object-based co-segmentation and reconstruction of scenes by enforcing consistent semantic labelling between views and over time. Semantic tracklets are introduced to enforce temporal coherence in semantic labelling and reconstruction between widely spaced instances of dynamic objects. Tracklets of dynamic objects enable unsupervised learning of appearance and shape priors that are exploited in joint segmentation and reconstruction. Evaluation on challenging indoor and outdoor sequences with hand-held moving cameras shows improved accuracy in segmentation, temporally coherent semantic labelling and 3D reconstruction of dynamic scenes.

    A Ahmed, A Hilton, F Mokhtarian (2003)Cyclification of Animation for Human Motion Synthesis, In: Eurographics Short Paper
    X Shen, P Palmer, P McLauchlan, A Hilton (2000)Error Propagation from Camera Motion to Epipolar Constraint, In: British Machine Vision Conference
    N Gkalelis, H Kim, A Hilton, N Nikolaidis, I Pitas (2017)The i3DPost multi-view and 3D human action/interaction
    C Budd, P Huang, A Hilton (2011)Hierarchical shape matching for temporally consistent 3D video, In: Proceedings of International Conference on 3D Imaging, Modeling, Processing, Visualization and Transmissionpp. 172-179

    In this paper we present a novel approach for temporal alignment of reconstructed mesh sequences with non-rigid surfaces to obtain a consistent representation. We propose a hierarchical scheme for non-sequential matching of frames across the sequence using shape similarity. This gives a tree structure which represents the optimal path for alignment of each frame in the sequence to minimize the change in shape. Non-rigid alignment is performed by recursively traversing the tree to align all frames. Non-sequential alignment reduces problems of drift or tracking failure which occur in previous sequential frame-to-frame techniques. Comparative evaluation on challenging 3D video sequences demonstrates that the proposed approach produces a temporally coherent representation with reduced error in shape and correspondence.

    H Kim, Jean-Yves Guillemaut, T Takai, M Sarim, A Hilton (2012)Outdoor Dynamic 3D Scene Reconstruction, In: H Gharavi (eds.), IEEE Transactions on Circuits and Systems for Video Technology22(11)pp. 1611-1622 IEEE

    Existing systems for 3D reconstruction from multiple view video use controlled indoor environments with uniform illumination and backgrounds to allow accurate segmentation of dynamic foreground objects. In this paper we present a portable system for 3D reconstruction of dynamic outdoor scenes which require relatively large capture volumes with complex backgrounds and non-uniform illumination. This is motivated by the demand for 3D reconstruction of natural outdoor scenes to support film and broadcast production. Limitations of existing multiple view 3D reconstruction techniques for use in outdoor scenes are identified. Outdoor 3D scene reconstruction is performed in three stages: (1) 3D background scene modelling using spherical stereo image capture; (2) multiple view segmentation of dynamic foreground objects by simultaneous video matting across multiple views; and (3) robust 3D foreground reconstruction and multiple view segmentation refinement in the presence of segmentation and calibration errors. Evaluation is performed on several outdoor productions with complex dynamic scenes including people and animals. Results demonstrate that the proposed approach overcomes limitations of previous indoor multiple view reconstruction approaches enabling high-quality free-viewpoint rendering and 3D reference models for production.

    Marco Volino, Peng Huang, Adrian Hilton (2018)Online interactive 4D character animation, In: Proceedings of the 20th International Conference on 3D Web Technology - Web3D '15pp. 289-295

    This paper presents a framework for creating realistic virtual characters that can be delivered via the Internet and interactively controlled in a WebGL enabled web-browser. Four-dimensional performance capture is used to capture realistic human motion and appearance. The captured data is processed into efficient and compact representations for geometry and texture. Motions are analysed against a high-level, user-defined motion graph and suitable inter- and intra-motion transitions are identified. This processed data is stored on a webserver and downloaded by a client application when required. A Javascript-based character animation engine is used to manage the state of the character which responds to user input and sends required frames to a WebGL-based renderer for display. Through the efficient geometry, texture and motion graph representations, a game character capable of performing a range of motions can be represented in 40-50 MB of data. This highlights the potential use of four-dimensional performance capture for creating web-based content. Datasets are made available for further research and an online demo is provided.

    M Sarim, A Hilton, J-Y Guillemaut (2011)TEMPORAL TRIMAP PROPAGATION FOR VIDEO MATTING USING INFERENTIAL STATISTICS, In: 2011 18TH IEEE INTERNATIONAL CONFERENCE ON IMAGE PROCESSING (ICIP)pp. 1745-1748
    Armin Mustafa, Marco Volino, Hansung Kim, Jean-Yves Guillemaut, Adrian Hilton (2020)Temporally coherent general dynamic scene reconstruction, In: International Journal of Computer Vision Springer

    Existing techniques for dynamic scene re- construction from multiple wide-baseline cameras pri- marily focus on reconstruction in controlled environ- ments, with fixed calibrated cameras and strong prior constraints. This paper introduces a general approach to obtain a 4D representation of complex dynamic scenes from multi-view wide-baseline static or moving cam- eras without prior knowledge of the scene structure, ap- pearance, or illumination. Contributions of the work are: An automatic method for initial coarse reconstruc- tion to initialize joint estimation; Sparse-to-dense tem- poral correspondence integrated with joint multi-view segmentation and reconstruction to introduce tempo- ral coherence; and a general robust approach for joint segmentation refinement and dense reconstruction of dynamic scenes by introducing shape constraint. Com- parison with state-of-the-art approaches on a variety of complex indoor and outdoor scenes, demonstrates im- proved accuracy in both multi-view segmentation and dense reconstruction. This paper demonstrates unsuper- vised reconstruction of complete temporally coherent 4D scene models with improved non-rigid object seg- mentation and shape reconstruction and its application to various applications such as free-view rendering and virtual reality.

    A Turkmani, A Hilton (2006)Appearane-Based Inner-Lip Detection, In: IET European Conference on Visual Media Productionpp. 176-176
    J Starck, A Hilton (2008)Model-based human shape reconstruction from multiple views, In: Computer Vision and Image Understanding111(2)pp. 179-194 Elsevier

    Image-based modelling allows the reconstruction of highly realistic digital models from real-world objects. This paper presents a model-based approach to recover animated models of people from multipleview video images. Two contributions are made, a multiple resolution model-based framework is introduced that combines multiple visual cues in reconstruction. Second, a novel mesh parameterisation is presented to preserve the vertex parameterisation in the model for animation. A prior humanoid surface model is first decomposed into multiple levels of detail and represented as a hierarchical deformable model for image fitting. A novel mesh parameterisation is presented that allows propagation of deformation in the model hierarchy and regularisation of surface deformation to preserve vertex parameterisation and animation structure. The hierarchical model is then used to fuse multipleshape cues from silhouette, stereo and sparse feature data in a coarse-to-fine strategy to recover a model that reproduces the appearance in the images. The framework is compared to physics-based deformable surface fitting at a single resolution, demonstrating an improved reconstruction accuracy against ground-truth data with a reduced model distortion. Results demonstrate realistic modelling of real people with accurate shape and appearance while preserving model structure for use in animation.

    A Hilton, J Starck, G Collins (2002)From 3D Shape Capture to Animated Models, In: IEEE Conference on 3D Data Processing, Visualisation and Transmissionpp. 246-255

    This paper presents a framework for construction of animated models from captured surface shape of real objects. Algorithms are introduced to transform the captured surface shape into a layered model. The layered model comprises an articulation structure, generic control model and a displacement map to represent the high-resolution surface detail. Novel methods are presented for automatic control model generation, shape constrained fitting and displacement mapping of the captured data. Results are demonstrated for surface shape captured using both multiple view images and active surface measurement. The framework enables rapid transformation of captured data into a structured representation suitable for realistic animation.

    M Sarim, A Hilton, Jean-Yves Guillemaut (2009)WIDE-BASELINE MATTE PROPAGATION FOR INDOOR SCENES, In: 2009 CONFERENCE FOR VISUAL MEDIA PRODUCTION: CVMP 2009pp. 195-204

    This paper presents a method to estimate alpha mattes for video sequences of the same foreground scene from wide-baseline views given sparse key-frame trimaps in a single view. A statistical inference framework is introduced for spatio-temporal propagation of high-confidence trimap labels between video sequences without a requirement for correspondence or camera calibration and motion estimation. Multiple view trimap propagation integrates appearance information between views and over time to achieve robust labelling in the presence of shadows, changes in appearance with view point and overlap between foreground and background appearance. Results demonstrate that trimaps are sufficiently accurate to allow high-quality video matting using existing single view natural image matting algorithms. Quantitative evaluation against ground-truth demonstrates that the approach achieves accurate matte estimation for camera views separated by up to 180◦ , with the same amount of manual interaction required for conventional single view video matting.

    Armin Mustafa, Marco Volino, Jean-Yves Guillemaut, Adrian Hilton (2018)4D Temporally Coherent Light-field Video, In: 3DV 2017 Proceedings IEEE

    Light-field video has recently been used in virtual and augmented reality applications to increase realism and immersion. However, existing light-field methods are generally limited to static scenes due to the requirement to acquire a dense scene representation. The large amount of data and the absence of methods to infer temporal coherence pose major challenges in storage, compression and editing compared to conventional video. In this paper, we propose the first method to extract a spatio-temporally coherent light-field video representation. A novel method to obtain Epipolar Plane Images (EPIs) from a spare lightfield camera array is proposed. EPIs are used to constrain scene flow estimation to obtain 4D temporally coherent representations of dynamic light-fields. Temporal coherence is achieved on a variety of light-field datasets. Evaluation of the proposed light-field scene flow against existing multiview dense correspondence approaches demonstrates a significant improvement in accuracy of temporal coherence.

    G Collins, A Hilton (2005)A Rigid Transform Basis for Animation Compression and Level of Detail, In: IMA Conference on Vision, Video and Graphicspp. 21-28

    We present a scheme for achieving level of detail and compression for animation sequences with known constant connectivity. We suggest compression is useful to automatically create low levels of detail in animations which may be more compressed than the original animation parameters and for high levels of detail where the original animation is expensive to compute. Our scheme is based on spatial segmentation of a base mesh into rigidly transforming segments and then temporal aggregation of these transformations. The result will approximate the given animation within a user specified tolerance which can be adjusted to give the required level of detail. A spatio-temporal smoothing algorithm is used on decoding to give acceptable animations. We show that the rigid transformation basis will span the space of all animations. We also show that the algorithm will converge to the specified tolerance. The algorithm is applied to several examples of synthetic animation and rate distortion curves are given which show that in some cases, the scheme outperforms current compressors.

    M Tejera, A Hilton (2013)Learning Part-Based Models for Animation from Surface Motion Capture, In: 2013 INTERNATIONAL CONFERENCE ON 3D VISION (3DV 2013)pp. 159-166 IEEE
    A Hilton, JB Roberts, O Hadded (1991)Autocorrelation Based Analysis of LDA Engine Data for Bias-Free Turbulence Estaimates, In: Society of Automotive Engineers International Congress
    James Woodcock, Jon Franombe, Andreas Franck, Philip Coleman, Richard Hughes, Hansung Kim, Qingju Liu, Dylan Menzies, Marcos F Simón Gálvez, Yan Tang, Tim Brookes, William J Davies, Bruno M Fazenda, Russell Mason, Trevor J Cox, Filippo Maria Fazi, Philip Jackson, Chris Pike, Adrian Hilton (2018)A Framework for Intelligent Metadata Adaptation in Object-Based Audio, In: AES E-Librarypp. P11-3 Audio Engineering Society

    Object-based audio can be used to customize, personalize, and optimize audio reproduction depending on the speci?c listening scenario. To investigate and exploit the bene?ts of object-based audio, a framework for intelligent metadata adaptation was developed. The framework uses detailed semantic metadata that describes the audio objects, the loudspeakers, and the room. It features an extensible software tool for real-time metadata adaptation that can incorporate knowledge derived from perceptual tests and/or feedback from perceptual meters to drive adaptation and facilitate optimal rendering. One use case for the system is demonstrated through a rule-set (derived from perceptual tests with experienced mix engineers) for automatic adaptation of object levels and positions when rendering 3D content to two- and ?ve-channel systems.

    M Klaudiny, M Tejera, C Malleson, J-Y Guillemaut, A Hilton (2020)SCENE Digital Cinema Datasets University of Surrey
    N Nadtoka, JR Tena, A Hilton, J Edge (2007)High-resolution Animation of Facial Dynamics, In: IET European Conference on Visual Media Production

    This paper presents a framework for performance-based animation and retargeting of high-resolution face models from motion capture. A novel method is introduced for learning a mapping between sparse 3D motion capture markers and dense high-resolution 3D scans of face shape and appearance. A high-resolution facial expression space is learnt from a set of 3D face scans as a person specific morphable model. Sparse 3D face points sampled at the motion capture marker positions are used to build a corresponding low-resolution expression space to represent the facial dynamics from motion capture. Radial basis function interpolation is used to automatically map the low-resolution motion capture of facial dynamics to the high-resolution facial expression space. This produces a high-resolution facial animation with the detailed shape and appearance of real facial dynamics. Retargeting is introduced to transfer facial expressions to a novel subject captured from a single photograph or 3D scan. The subject specific high- resolution expression space is mapped to the novel subject based on anatomical differences in face shape. Results facial animation and retargeting demonstrate realistic animation of expressions from motion capture.

    Graham Miller, Adrian Hilton (2007)Safe Hulls, In: IET European Conference on Visual Media Production

    The visual hull is widely used as a proxy for novel view synthesis in computer vision. This paper introduces the safe hull, the first visual hull reconstruction technique to produce a surface containing only foreground parts. A theoretical basis underlies this novel approach which, unlike any previous work, can also identify phantom volumes attached to real objects. Using an image-based method, the visual hull is constructed with respect to each real view and used to identify safe zones in the original silhouettes. The safe zones define volumes known to only contain surface corresponding to a real object. The zones are used in a second reconstruction step to produce a surface without phantom volumes. Results demonstrate the effectiveness of this method for improving surface shape and scene realism, and its advantages over heuristic techniques.

    H Kim, A Hilton (2015)Block world reconstruction from spherical stereo image pairs, In: Computer Vision and Image Understanding139pp. 104-121 ELSEVIER

    We propose a block-based scene reconstruction method using multiple stereo pairs of spherical images. We assume that the urban scene consists of axis-aligned planar structures (Manhattan world). Captured spherical stereo images are converted into six central-point perspective images by cubic projection and fa____c cade alignment. Depth information is recovered by stereo matching between images. Semantic regions are segmented based on colour, edge and normal information. Independent 3D rectangular planes are constructed by fitting planes aligned with the principal axes of the segmented 3D points. Finally cuboid-based scene structure is recovered from multiple viewpoints by merging and refining planes based on connectivity and visibility. The reconstructed model efficiently shows the structure of the scene with a small amount of data.

    Qingju Liu, T deCampos, Wenwu Wang, Adrian Hilton (2016)Identity association using PHD filters in multiple head tracking with depth sensors, In: IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP 2016) Proceedings

    The work on 3D human pose estimation has seen a significant amount of progress in recent years, particularly due to the widespread availability of commodity depth sensors. However, most pose estimation methods follow a tracking-as-detection approach which does not explicitly handle occlusions, thus introducing outliers and identity association issues when multiple targets are involved. To address these issues, we propose a new method based on Probability Hypothesis Density (PHD) filter. In this method, the PHD filter with a novel clutter intensity model is used to remove outliers in the 3D head detection results, followed by an identity association scheme with occlusion detection for the targets. Experimental results show that our proposed method greatly mitigates the outliers, and correctly associates identities to individual detections with low computational cost.

    J Edge, A Hilton (2006)Visual Speech Synthesis from 3D Video, In: IET European Conference on Visual Media Productionpp. 174-174

    In this paper we describe a parameterisation of lip movements which maintains the dynamic structure inherent in the task of producing speech sounds. A stereo capture system is used to reconstruct 3D models of a speaker producing sentences from the TIMIT corpus. This data is mapped into a space which maintains the relationships between samples and their temporal derivatives. By incorporating dynamic information within the parameterisation of lip movements we can model the cyclical structure, as well as the causal nature of speech movements as described by an underlying visual speech manifold. It is believed that such a structure will be appropriate to various areas of speech modeling, in particular the synthesis of speech lip movements.

    Hansung Kim, Luca Remaggi, Philip Jackson, Adrian Hilton (2019)Spatial Audio Reproduction System for VR Using 360 degree Cameras, In: The 12th Asia Pacific Workshop on Mixed and Augmented Reality (APMAR 2019) Proceedings

    In order to maximise the immersion in VR environments, a plausible spatial audio reproduction synchronised with visual information is essential. In this work, we propose a pipeline to create plausible interactive audio from a pair of 360 degree cameras.

    ADM Hilton, Jean-Yves Guillemaut, JJ Kilner, O Grau, G Thomas (2011)3D-TV Production from Conventional Cameras for Sports Broadcast, In: IEEE Transactions Broadcasting57(2)pp. 462-476 IEEE

    3DTV production of live sports events presents a challenging problem involving conflicting requirements of main- taining broadcast stereo picture quality with practical problems in developing robust systems for cost effective deployment. In this paper we propose an alternative approach to stereo production in sports events using the conventional monocular broadcast cameras for 3D reconstruction of the event and subsequent stereo rendering. This approach has the potential advantage over stereo camera rigs of recovering full scene depth, allowing inter-ocular distance and convergence to be adapted according to the requirements of the target display and enabling stereo coverage from both existing and ‘virtual’ camera positions without additional cameras. A prototype system is presented with results of sports TV production trials for rendering of stereo and free-viewpoint video sequences of soccer and rugby.

    Conventional stereoscopic video content production requires use of dedicated stereo camera rigs which is both costly and lacking video editing flexibility. In this paper, we propose a novel approach which only requires a small number of standard cameras sparsely located around a scene to automatically convert the monocular inputs into stereoscopic streams. The approach combines a probabilistic spatio-temporal segmentation framework with a state-of-the-art multi-view graph-cut reconstruction algorithm, thus providing full control of the stereoscopic settings at render time. Results with studio sequences of complex human motion demonstrate the suitability of the method for high quality stereoscopic content generation with minimum user interaction.

    Luca Remaggi, Philip J. B. Jackson, Hansung Kim, Adrian Hilton (2018)An Audio-Visual Method for Room Boundary Estimation and Material Recognition, In: Proceedings of 2018 Workshop on Audio-Visual Scene Understanding for Immersive Multimedia, Seoul, Republic of Korea, October 26, 2018 (AVSU’18)pp. 3-9 Association for Computing Machinery (ACM)

    In applications such as virtual and augmented reality, a plausible and coherent audio-visual reproduction can be achieved by deeply understanding the reference scene acoustics. This requires knowledge of the scene geometry and related materials. In this paper, we present an audio-visual approach for acoustic scene understanding. We propose a novel material recognition algorithm, that exploits information carried by acoustic signals. The acoustic absorption coefficients are selected as features. The training dataset was constructed by combining information available in the literature, and additional labeled data that we recorded in a small room having short reverberation time (RT60). Classic machine learning methods are used to validate the model, by employing data recorded in five rooms, having different sizes and RT60s. The estimated materials are utilized to label room boundaries, reconstructed by a visionbased method. Results show 89 % and 80 % agreement between the estimated and reference room volumes and materials, respectively.

    Y Li, A Hilton, J Illingworth (2012)Towards Reliable Real-Time Multiview Tracking, In: IEEE International Workshop on Multiple Object Trackingpp. 43-50

    We address the problem of reliable real-time 3D-tracking of multiple objects which are observed in multiple wide-baseline camera views. Establishing the spatio-temporal correspondence is a problem with combinatorial complexity in the number of objects and views. In addition vision based tracking suffers from the ambiguities introduced by occlusion, clutter and irregular 3D motion. We present a discrete relaxation algorithm for reducing the intrinsic combinatorial complexity by pruning the decision tree based on unreliable prior information from independent 2D-tracking for each view. The algorithm improves the reliability of spatio-temporal correspondence by simultaneous optimisation over multiple views in the case where 2D-tracking in one or more views is ambiguous. Application to the 3D reconstruction of human movement, based on tracking of skin-coloured regions in three views, demonstrates considerable improvement in reliability and performance. The results demonstrate that the optimisation over multiple views gives correct 3D reconstruction and object labeling in the presence of incorrect 2D-tracking whilst maintaining real-time performance

    A Mustafa, H Kim, J-Y Guillemaut, ADM Hilton (2016)Temporally coherent 4D reconstruction of complex dynamic scenes, In: CVPR 2016 Proceedings

    This paper presents an approach for reconstruction of 4D temporally coherent models of complex dynamic scenes. No prior knowledge is required of scene structure or camera calibration allowing reconstruction from multiple moving cameras. Sparse-to-dense temporal correspondence is integrated with joint multi-view segmentation and reconstruction to obtain a complete 4D representation of static and dynamic objects. Temporal coherence is exploited to overcome visual ambiguities resulting in improved reconstruction of complex scenes. Robust joint segmentation and reconstruction of dynamic objects is achieved by introducing a geodesic star convexity constraint. Comparative evaluation is performed on a variety of unstructured indoor and outdoor dynamic scenes with hand-held cameras and multiple people. This demonstrates reconstruction of complete temporally coherent 4D scene models with improved nonrigid object segmentation and shape reconstruction.

    Yang Liu, Wenwu Wang, Jonathon Chambers, Volkan Kilic, Adrian Hilton (2017)Particle flow SMC-PHD filter for audio-visual multi-speaker tracking. Proc. 13th International Conference on Latent Variable Analysis and Signal Separation(LVA/ICA 2017), Grenoble, France, February 21-23, 2017., In: P Tichavský, M Babaie-Zadeh, O Michel, N Thirion-Moreau (eds.), Latent Variable Analysis and Signal Separation. LVA/ICA 201710169pp. 344-353 Springer

    Sequential Monte Carlo probability hypothesis density (SMC- PHD) filtering has been recently exploited for audio-visual (AV) based tracking of multiple speakers, where audio data are used to inform the particle distribution and propagation in the visual SMC-PHD filter. However, the performance of the AV-SMC-PHD filter can be affected by the mismatch between the proposal and the posterior distribution. In this paper, we present a new method to improve the particle distribution where audio information (i.e. DOA angles derived from microphone array measurements) is used to detect new born particles and visual information (i.e. histograms) is used to modify the particles with particle flow (PF). Using particle flow has the benefit of migrating particles smoothly from the prior to the posterior distribution. We compare the proposed algorithm with the baseline AV-SMC-PHD algorithm using experiments on the AV16.3 dataset with multi-speaker sequences.

    J Mitchelson, A Hilton (2002)Wand-based Calibration of Multiple Cameras, In: British Machine Vision Association workshop on Multiple Views
    Qingju Liu, T deCampos, Wenwu Wang, Philip Jackson, Adrian Hilton (2016)Person tracking using audio and depth cues, In: International Conference on Computer Vision (ICCV) Workshop on 3D Reconstruction and Understanding with Video and Soundpp. 709-717

    In this paper, a novel probabilistic Bayesian tracking scheme is proposed and applied to bimodal measurements consisting of tracking results from the depth sensor and audio recordings collected using binaural microphones. We use random finite sets to cope with varying number of tracking targets. A measurement-driven birth process is integrated to quickly localize any emerging person. A new bimodal fusion method that prioritizes the most confident modality is employed. The approach was tested on real room recordings and experimental results show that the proposed combination of audio and depth outperforms individual modalities, particularly when there are multiple people talking simultaneously and when occlusions are frequent.

    A HILTON, J ILLINGWORTH, T WINDEATT (1994)STATISTICS OF SURFACE CURVATURE ESTIMATES, In: PROCEEDINGS OF THE 12TH IAPR INTERNATIONAL CONFERENCE ON PATTERN RECOGNITION - CONFERENCE A: COMPUTER VISION & IMAGE PROCESSINGpp. 37-41
    AJ Stoddart, A Hilton, J Illingworth (1994)Slime: A new deformable surface, In: BMVC
    JB Roberts, ADM Hilton (2001)A direct transform method for the analysis of laser Doppler anemometry engine data, In: PROCEEDINGS OF THE INSTITUTION OF MECHANICAL ENGINEERS PART D-JOURNAL OF AUTOMOBILE ENGINEERING215(D6)pp. 725-738 PROFESSIONAL ENGINEERING PUBLISHING LTD
    A Hilton, JB Roberts, O Hadded (1992)Comparative Evaluation of Techniques for Estimating Turbulent Flow Parameters from In-Cylinder LDA Engine Data, In: Fifth International Symposium on Applications of Laser Anemometry to Fluid Mechanics, Lisbon, Portugalpp. 130-138
    H Kim, A Hilton (2008)Region-based Foreground Extraction, In: European Conference on Visual Media Production

    We propose a region-based method to extract foreground regions from colour video sequences. The foreground region is decided by voting with scores from background subtraction to the sub-regions by graph- based segmentation. Experiments show that the proposed algorithm improves on conventional approaches especially in strong shadow regions.

    P Stroia-Williams, A Hilton (2008)Example-based Reflectance Estimation for Capturing Relightable Models of People, In: European Conference on Visual Media Productionpp. 1-10

    We present a new approach to reflectance estimation for dynamic scenes. Non-parametric image statistics are used to transfer reflectance properties from a static example set to a dynamic image sequence. The approach allows reflectance estimation for surface materials with inhomogeneous appearance, such as those which commonly occur with patterned or textured clothing. Material reflectance properties are initially estimated from static images of the subject under multiple directional illuminations using photometric stereo. The estimated reflectance together with the corresponding image under uniform ambient illumination form a prior set of reference material observations. Material reflectance properties are then estimated for video sequences of a moving person captured under uniform ambient illumination by matching the observed local image statistics to the reference observations. Results demonstrate that the transfer of reflectance properties enables estimation of the dynamic surface normals and subsequent relighting. This approach overcomes limitations of previous work on material transfer and relighting of dynamic scenes which was limited to surfaces with regions of homogeneous reflectance. We evaluate for relighting 3D model sequences reconstructed from multiple view video. Comparison to previous model relighting demonstrates improved reproduction of detailed texture and shape dynamics.

    P Huang, A Starck, A Hilton (2007)Temporal 3D Shape Matching, In: IET European Conference on Visual Media Productionpp. 1-8

    This paper introduces a novel 4D shape descriptor to match temporal surface sequences. A quantitative evaluation based on the receiver-operator characteristic (ROC) curve is presented to compare the performance of conventional 3D shape descriptors with and without using a time filter. Feature- based 3D shape descriptors including shape distribution (Osada et al., 2002 ), spin image (Johnson et al., 1999), shape histogram (Ankest et al., 1999) and spherical harmonics (Kazhdan et al., 2003) are considered. Evaluation shows that filtered descriptors outperform unfiltered descriptors and the best performing volume-sampling shape-histogram descriptor is extended to define a new 4D "shape-flow" descriptor. Shape-flow matching demonstrates improved performance in the context of matching time-varying sequences which is motivated by the requirement to connect similar sequences for animation production. Both simulated and real 3D human surface motion sequences are used for evaluation.

    A Hilton, JB Roberts, O Hadded (1991)Autocorrelation Based Analysis of Ensemble Averaged LDA Engine Data for Bias-Free Turbulence Estimates: A Unified Approach, In: Journal of the Society of Automotive Engineering SAE910479
    G Collins, A Hilton (2005)Spatio-Temporal Fusion of Multiple View Video Rate 3D Surfaces, In: Fifth International Conference on 3-D Digital Imaging and Modeling (3DIM’05)pp. 142-149

    We consider the problem of geometric integration and representation of multiple views of non-rigidly deforming 3D surface geometry captured at video rate. Instead of treating each frame as a separate mesh we present a representation which takes into consideration temporal and spatial coherence in the data where possible. We first segment gross base transformations using correspondence based on a closest point metric and represent these motions as piecewise rigid transformations. The remaining residual is encoded as displacement maps at each frame giving a displacement video. At both these stages occlusions and missing data are interpolated to give a representation which is continuous in space and time. We demonstrate the integration of multiple views for four different non-rigidly deforming scenes: hand, face, cloth and a composite scene. The approach achieves the integration of multiple-view data at different times into one representation which can processed and edited.

    C Budd, P Huang, M Klaudiny, A Hilton (2012)Global Non-rigid Alignment of Surface Sequences, In: International Journal of Computer Vision Springer Verlag

    This paper presents a general approach based on the shape similarity tree for non-sequential alignment across databases of multiple unstructured mesh sequences from non-rigid surface capture. The optimal shape similarity tree for non-rigid alignment is defined as the minimum spanning tree in shape similarity space. Non-sequential alignment based on the shape similarity tree minimises the total non-rigid deformation required to register all frames in a database into a consistent mesh structure with surfaces in correspondence. This allows alignment across multiple sequences of different motions, reduces drift in sequential alignment and is robust to rapid non-rigid motion. Evaluation is performed on three benchmark databases of 3D mesh sequences with a variety of complex human and cloth motion. Comparison with sequential alignment demonstrates reduced errors due to drift and improved robustness to large non-rigid deformation, together with global alignment across multiple sequences which is not possible with previous sequential approaches. © 2012 The Author(s).

    M Sarim, A Hilton, J Guillemaut (2009)Non-parametric patch based video matting

    In computer vision, matting is the process of accurate foreground estimation in images and videos. In this paper we presents a novel patch based approach to video matting relying on non-parametric statistics to represent image variations in appearance. This overcomes the limitation of parametric algorithms which only rely on strong colour correlation between the nearby pixels. Initially we construct a clean background by utilising the foreground object’s movement across the background. For a given frame, a trimap is constructed using the background and the last frame’s trimap. A patch-based approach is used to estimate the foreground colour for every unknown pixel and finally the alpha matte is extracted. Quantitative evaluation shows that the technique performs better, in terms of the accuracy and the required user interaction, than the current state-of-the-art parametric approaches.

    Hansung Kim, A Evans, J Blat, Adrian Hilton (2016)Multi-modal Visual Data Registration for Web-based Visualisation in Media Production, In: IEEE Transactions on Circuits and Systems for Video Technology28(4)pp. 863-877 IEEE

    Recent developments of video and sensing technology can lead to large amounts of digital media data. Current media production rely on both video from the principal camera together with a wide variety of heterogeneous source of supporting data (photos, LiDAR point clouds, witness video camera, HDRI and depth imagery). Registration of visual data acquired from various 2D and 3D sensing modalities is challenging because current matching and registration methods are not appropriate due to differences in formats and noise types of multi-modal data. A combined 2D/3D visualisation of this registered data allows an integrated overview of the entire dataset. For such a visualisation a web-based context presents several advantages. In this paper we propose a unified framework for registration and visualisation of this type of visual media data. A new feature description and matching method is proposed, adaptively considering local geometry, semi-global geometry and colour information in the scene for more robust registration. The resulting registered 2D/3D multi-modal visual data is too large to be downloaded and viewed directly via the web browser while maintaining an acceptable user experience. Thus, we employ hierarchical techniques for compression and restructuring to enable efficient transmission and visualisation over the web, leading to interactive visualisation as registered point clouds, 2D images, and videos in the browser, improving on the current state of the art techniques for web-based visualisation of big media data. This is the first unified 3D web-based visualisation of multi-modal visual media production datasets. The proposed pipeline is tested on big multimodal dataset typical of film and broadcast production which are made publicly available. The proposed feature description method shows two times higher precision of feature matching and more stable registration performance than existing 3D feature descriptors.

    M Klaudiny, A Hilton (2012)High-detail 3D capture and non-sequential alignment of facial performance, In: Proceedings - 2nd Joint 3DIM/3DPVT Conference: 3D Imaging, Modeling, Processing, Visualization and Transmission, 3DIMPVT 2012pp. 17-24

    This paper presents a novel system for the 3D capture of facial performance using standard video and lighting equipment. The mesh of an actor's face is tracked non-sequentially throughout a performance using multi-view image sequences. The minimum spanning tree calculated in expression dissimilarity space defines the traversal of the sequences optimal with respect to error accumulation. A robust patch-based frame-to-frame surface alignment combined with the optimal traversal significantly reduces drift compared to previous sequential techniques. Multi-path temporal fusion resolves inconsistencies between different alignment paths and yields a final mesh sequence which is temporally consistent. The surface tracking framework is coupled with photometric stereo using colour lights which captures metrically correct skin geometry. High-detail UV normal maps corrected for shadow and bias artefacts augment the temporally consistent mesh sequence. Evaluation on challenging performances by several actors demonstrates the acquisition of subtle skin dynamics and minimal drift over long sequences. A quantitative comparison to a state-of-the-art system shows similar quality of temporal alignment. © 2012 IEEE.

    Estephan Dazzi, Teofilo de Campos, Adrian Hilton, Roberto M. Cesar Jr. (2017)Scalable object instance recognition based on keygraph matching, In: Pattern Recognition Letters114pp. 53-62 Elsevier

    We propose a generalisation of the local feature matching framework, where keypoints are replaced by k-keygraphs, i.e., isomorphic directed attributed graphs of cardinality k whose vertices are keypoints. Keygraphs have structural and topological properties which are discriminative and efficient to compute, based on graph edge length and orientation as well as vertex scale and orientation. Keypoint matching is performed based on descriptor similarity. Next, 2-keygraphs are calculated; as a result, the number of incorrect keypoint matches reduced in 75% (while the correct keypoint matches were preserved). Then, 3-keygraphs are calculated, followed by 4-keygraphs; this yielded a significant reduction of 99% in the number of remaining incorrect keypoint matches. The stage that finds 2-keygraphs has a computational cost equal to a small fraction of the cost of the keypoint matching stage, while the stages that find 3-keygraphs or 4-keygraphs have a negligible cost. In the final stage, RANSAC finds object poses represented as affine transformations mapping images. Our experiments concern large-scale object instance recognition subject to occlusion, background clutter and appearance changes. By using 4-keygraphs, RANSAC needed 1% of the iterations in comparison with 2-keygraphs or simple keypoints. As a result, using 4-keygraphs provided a better efficiency as well as allowed a larger number of initial keypoints matches to be established, which increased performance.

    Hansung Kim, Adrian Hilton, Kwanghoon Sohn, Bumsub Ham, Changjae Oh (2019)OCEAN: Object-Centric Arranging Network for Self-supervised Visual Representations Learning, In: Expert Systems with Applications125pp. pp 281-292 Elsevier

    Learning visual representations plays an important role in computer vision and machine learning applications. It facilitates a model to understand and perform high-level tasks intelligently. A common approach for learning visual representations is supervised one which requires a huge amount of human annotations to train the model. This paper presents a self-supervised approach which learns visual representations from input images without human annotations. We learn the correct arrangement of object proposals to represent an image using a convolutional neural network (CNN) without any manual annotations. We hypothesize that the network trained for solving this problem requires the embedding of semantic visual representations. Unlike existing approaches that use uniformly sampled patches, we relate object proposals that contain prominent objects and object parts. More specifically, we discover the representation that considers overlap, inclusion, and exclusion relationship of proposals as well as their relative position. This allows focusing on potential objects and parts rather than on clutter. We demonstrate that our model outperforms existing self-supervised learning methods and can be used as a generic feature extractor by applying it to object detection, classification, action recognition, image retrieval, and semantic matching tasks.

    J Starck, G Miller, A Hilton (2005)Video-Based Character Animation, In: ACM SIGGRAPH/Eurographics Symposium on Computer Animation

    In this paper we introduce a video-based representation for free viewpoint visualization and motion control of 3D character models created from multiple view video sequences of real people. Previous approaches to videobased rendering provide no control of scene dynamics to manipulate, retarget, and create new 3D content from captured scenes. Here we contribute a new approach, combining image based reconstruction and video-based animation to allow controlled animation of people from captured multiple view video sequences. We represent a character as a motion graph of free viewpoint video motions for animation control. We introduce the use of geometry videos to represent reconstructed scenes of people for free viewpoint video rendering. We describe a novel spherical matching algorithm to derive global surface to surface correspondence in spherical geometry images for motion blending and the construction of seamless transitions between motion sequences. Finally, we demonstrate interactive video-based character animation with real-time rendering and free viewpoint visualization. This approach synthesizes highly realistic character animations with dynamic surface shape and appearance captured from multiple view video of people.

    A Hilton, G Godin, C Shu, T Masuda (2011)Special issue on 3D imaging and modelling, In: COMPUTER VISION AND IMAGE UNDERSTANDING115(5)pp. 559-560 ACADEMIC PRESS INC ELSEVIER SCIENCE
    J Kilner, J Starck, A Hilton, O Grau (2007)Dual-mode deformable models for free-viewpoint video of sports events, In: G Godin, P Hebert, T Masuda, G Taubin (eds.), 3DIM 2007: Sixth International Conference on 3-D Digital Imaging and Modeling, Proceedingspp. 177-184
    Andrew Gilbert, Marco Volino, John Collomosse, Adrian Hilton (2018)Volumetric performance capture from minimal camera viewpoints, In: V Ferrari, M Hebert, C Sminchisescu, Y Weiss (eds.), Computer Vision – ECCV 2018. ECCV 2018. Lecture Notes in Computer Science11215pp. 591-607 Springer Science+Business Media

    We present a convolutional autoencoder that enables high fidelity volumetric reconstructions of human performance to be captured from multi-view video comprising only a small set of camera views. Our method yields similar end-to-end reconstruction error to that of a prob- abilistic visual hull computed using significantly more (double or more) viewpoints. We use a deep prior implicitly learned by the autoencoder trained over a dataset of view-ablated multi-view video footage of a wide range of subjects and actions. This opens up the possibility of high-end volumetric performance capture in on-set and prosumer scenarios where time or cost prohibit a high witness camera count.

    A Ahmed, A Hilton, F Mokhtarian (2004)Intuitive Parametric Synthesis of Human Animation Sequences, In: IEEE Computer Animation and Social Agents
    TB Moeslund, A Hilton, V Krüger, L Sigal (2011)Visual Analysis of Humans: Looking at People Springer-Verlag New York Inc
    Charles Malleson, Jean-Yves Guillemaut, Adrian Hilton (2018)Hybrid modelling of non-rigid scenes from RGBD cameras, In: IEEE Transactions on Circuits and Systems for Video Technology IEEE

    Recent advances in sensor technology have introduced low-cost RGB video plus depth sensors, such as the Kinect, which enable simultaneous acquisition of colour and depth images at video rates. This paper introduces a framework for representation of general dynamic scenes from video plus depth acquisition. A hybrid representation is proposed which combines the advantages of prior surfel graph surface segmentation and modelling work with the higher-resolution surface reconstruction capability of volumetric fusion techniques. The contributions are (1) extension of a prior piecewise surfel graph modelling approach for improved accuracy and completeness, (2) combination of this surfel graph modelling with TSDF surface fusion to generate dense geometry, and (3) proposal of means for validation of the reconstructed 4D scene model against the input data and efficient storage of any unmodelled regions via residual depth maps. The approach allows arbitrary dynamic scenes to be efficiently represented with temporally consistent structure and enhanced levels of detail and completeness where possible, but gracefully falls back to raw measurements where no structure can be inferred. The representation is shown to facilitate creative manipulation of real scene data which would previously require more complex capture setups or manual processing.

    S Pabst, H Kim, L Polok, V Ila, T Waine, A Hilton, J Clifford (2015)Jigsaw - Multi-Modal Big Data Management in Digital Film Production
    Philip Coleman, Andreas Franck, Jon Francombe, Qingju Liu, Teofilo de Campos, Richard Hughes, Dylan Menzies, Marcos Simón Gálvez, Yan Tang, James Woodcock, Frank Melchior, Chris Pike, Filippo Fazi, Trevor Cox, Adrian Hilton, PHILIP J B JACKSON (2020)S3A Audio-Visual System for Object-Based Audio University of Surrey
    JB Roberts, A Hilton (2001)A Direct Transform Method for the Analysis of LDA Engine Data, In: I.Mech.E. Journal of Automotive Engineering251D
    Armin Mustafa, Adrian Hilton (2019)Understanding real-world scenes for human-like machine perception, In: Proceedings of the Machine Intelligence 21 (MI21-HLC) workshop Imperial College Press

    The rise of autonomous machines in our day-to-day lives has led to an increasing demand for machine perception of real-world to be more robust, accurate and human-like. The research in visual scene un- derstanding over the past two decades has focused on machine perception in controlled environments such as indoor, static and rigid objects. There is a gap in literature for machine perception in general complex scenes (outdoor with multiple interacting people). The proposed research ad- dresses the limitations of existing methods by proposing an unsupervised framework to simultaneously model, semantically segment and estimate motion for general dynamic scenes captured from multiple view videos with a network of static or moving cameras. In this talk I will explain the proposed joint framework to understand general dynamic scenes for ma- chine perception; give a comprehensive performance evaluation against state-of-the-art techniques on challenging indoor and outdoor sequences; and demonstrate applications such as virtual, augmented, mixed reality (VR/AR/MR) and broadcast production (Free-view point video - FVV).

    Luca Remaggi, Hansung Kim, Philip Jackson, Filippo Maria Fazi, Adrian Hilton (2018)Acoustic reflector localization and classification, In: Proceedings of ICASSP 2018 - 2018 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) Institute of Electrical and Electronics Engineers (IEEE)

    The process of understanding acoustic properties of environments is important for several applications, such as spatial audio, augmented reality and source separation. In this paper, multichannel room impulse responses are recorded and transformed into their direction of arrival (DOA)-time domain, by employing a superdirective beamformer. This domain can be represented as a 2D image. Hence, a novel image processing method is proposed to analyze the DOA-time domain, and estimate the reflection times of arrival and DOAs. The main acoustically reflective objects are then localized. Recent studies in acoustic reflector localization usually assume the room to be free from furniture. Here, by analyzing the scattered reflections, an algorithm is also proposed to binary classify reflectors into room boundaries and interior furniture. Experiments were conducted in four rooms. The classification algorithm showed high quality performance, also improving the localization accuracy, for non-static listener scenarios.

    A Hilton, J Illingworth (2000)Geometric Fusion for a Hand-held 3D Sensor, In: Machine Vision and Applications12(1)pp. 44-51 Springer
    A Hilton, AJ Stoddart, J Illingworth, T Windeatt (1996)Reliable Surface Reconstruction from Multiple Range Images, In: 4th European Conference on Computer Vision1064

    This paper addresses the problem of reconstructing an integrated 3D model from multiple 2.5D range images. A novel integration algorithm is presented based on a continuous implicit surface representation. This is the first reconstruction algorithm to use operations in 3D space only. The algorithm is guaranteed to reconstruct the correct topology of surface features larger than the range image sampling resolution. Reconstruction of triangulated models from multi-image data sets is demonstrated for complex objects. Performance characterization of existing range image integration algorithms is addressed in the second part of this paper. This comparison defines the relative computational complexity and geometric limitations of existing integration algorithms.

    P McLauchlan, X Shen, P Palmer, A Manessis, A Hilton (2000)Surface-Based Structure-from-Motion using Feature Groupings, In: IEEE International Asian Conference on Computer Vision
    A Hilton, M Kalkavouras, G Collins (2004)MELIES: 3D Studio Production of Animated Actor Models, In: IEE European Conference on Visual Media Production
    Davide Berghi, Hanne Stenzel, Marco Volino, Adrian Hilton, Philip Jackson (2020)Audio-Visual Spatial Alignment Requirements of Central and Peripheral Object Events, In: IEEE VR 2020 IEEE

    Immersive audio-visual perception relies on the spatial integration of both auditory and visual information which are heterogeneous sensing modalities with different fields of reception and spatial resolution. This study investigates the perceived coherence of audiovisual object events presented either centrally or peripherally with horizontally aligned/misaligned sound. Various object events were selected to represent three acoustic feature classes. Subjective test results in a simulated virtual environment from 18 participants indicate a wider capture region in the periphery, with an outward bias favoring more lateral sounds. Centered stimulus results support previous findings for simpler scenes.

    JR Tena, M Hamouz, A Hilton, J Illingworth (2006)A Validation Method for Dense Non-rigid 3D Face Registration, In: IEEE Conf. on Advanced Video and Signal-based Surveillance
    Matthew Trumble, Andrew Gilbert, Adrian Hilton, John Collomosse (2016)Deep Convolutional Networks for Marker-less Human Pose Estimation from Multiple Views, In: Proceedings of CVMP 2016. The 13th European Conference on Visual Media Production

    We propose a human performance capture system employing convolutional neural networks (CNN) to estimate human pose from a volumetric representation of a performer derived from multiple view-point video (MVV).We compare direct CNN pose regression to the performance of an affine invariant pose descriptor learned by a CNN through a classification task. A non-linear manifold embedding is learned between the descriptor and articulated pose spaces, enabling regression of pose from the source MVV. The results are evaluated against ground truth pose data captured using a Vicon marker-based system and demonstrate good generalisation over a range of human poses, providing a system that requires no special suit to be worn by the performer.

    Andrew Gilbert, Matt Trumble, Adrian Hilton, John Collomosse (2018)Inpainting of Wide-baseline Multiple Viewpoint Video, In: IEEE Transactions on Visualization and Computer Graphics Institute of Electrical and Electronics Engineers (IEEE)

    We describe a non-parametric algorithm for multiple-viewpoint video inpainting. Uniquely, our algorithm addresses the domain of wide baseline multiple-viewpoint video (MVV) with no temporal look-ahead in near real time speed. A Dictionary of Patches (DoP) is built using multi-resolution texture patches reprojected from geometric proxies available in the alternate views. We dynamically update the DoP over time, and a Markov Random Field optimisation over depth and appearance is used to resolve and align a selection of multiple candidates for a given patch, this ensures the inpainting of large regions in a plausible manner conserving both spatial and temporal coherence. We demonstrate the removal of large objects (e.g. people) on challenging indoor and outdoor MVV exhibiting cluttered, dynamic backgrounds and moving cameras.