I-Lab: Media adaptation

Multimedia content is accessed, shared and distributed by users across devices, platforms and networks. Media adaption strategies are necessary for universal access and delivery of digital media content; our research focuses on providing technologies to adapt media content.

Effective strategies are needed for transparent access and delivery of media content universally to address growing heterogeneity in media device capabilities, access network characteristics, content representation formats and standards, usage environment properties, users’ preferences, etc.

Media adaptation provides the necessary technologies for enabling such strategies, as it is the process of transforming input media content to output media in order to meet diverse network, terminal, environment, user, etc requirements to increase the usability of media content.

I-Lab media adaptation research focuses on providing such technologies to adapt multimedia content, including relevant adaptation decision taking methods to accompany those adaptation operations for both 2D and 3D digital media that are accessed, shared, and distributed in an end-to-end digital media delivery chain.

In a heterogeneous media networking environment, the objectives of media adaptation are to:

  • Enhance rate-distortion performance of digital media and improve overall user experience during delivery/access
  • Boost usability and effectiveness of content scalability
  • Provide error robustness in adverse conditions
  • Increase interoperability and accessibility of content
  • Provide universal accessibility to protected and digital rights managed (DRM-enabled) content.

Adaptation of media is performed in two complementary stages: the first is dedicated for taking decisions in response to adaptation requests, and the second is the execution of necessary adaptation algorithms in line with the decisions taken. The former is performed by a module traditionally referred to as an adaptation decision engine (ADE) while the latter comprises the actual adaptation operations conducted on the media content by an adaptation engine (AE).

The research conducted in the I-Lab addresses both stages of adaptation to provide solutions for re-purposing media content through transrating, transcoding, spatio-temporal scaling, error resilience insertion, view selection, region-of-interest (RoI) and visual attention modelling based adaptation, as well as adapting secured content transparently. Recent work has also targeted investigations into ambient illumination as an effective context in the usage environment, and developing models and adaptation strategies using this context for enhancement of users’ viewing experiences for both 2D and 3D visual material.

Research interests and expertise

  1. MPEG-21 digital item adaptation (DIA), context awareness and quality of viewing experience in media communications
  2. RoI and visual attention modelling based video adaptation
  3. Error robustness adaptation through prioritising 2D/3D video content and error resilience insertion
  4. Content and context aware scalable video adaptation
  5. Adaptation aware encryption of scalable video for secure content adaptation.

Additional research themes

  • Quality of Experience (QoE) based digital media content adaptation
  • 2D/3D video adaptation considering new environmental context, such as ambient illumination (for more details and results, please see the relevant research item under "QoE research @I-Lab")
  • Audio assisted video adaptation and vice versa
  • Multi-view scalable video adaptation (e.g. view and viewpoint selection based adaptation)
  • Related adaptation decision taking strategies.

As it provides a seamless universal access to content from anywhere anytime and using any device, the application areas of media adaptation thus are diverse, such as:

  • User centric content access, sharing, and consumption over heterogeneous media delivery platforms
  • Social media networking, multi-party remote collaboration, conferencing, collaborative design, education, training, etc
  • e-Health, e-Learning, e-Culture, e-Tourism, e-Government, etc
  • Media retrieval, personalised multimedia services, media content analysis, next generation networking, QoE-driven media delivery, surveillance, etc.

In addition to postgraduate level research pursued at the I-Lab, the aforementioned diverse work items in the context of media adaptation research have been conducted within various EU framework programme projects, namely VISNET I and II NoEs, DIOMEDES STREP, MUSCADE and ROMEO IP projects, and have been reported in prestigious journals and leading international conferences with very positive feedback from the relevant research communities.

The relevant list of publications indicates the diversity and quality of the research outputs produced by the I-Lab researchers working in the media adaptation research topic.

For more detailed information on this topic, please contact:

Contact us

Find us

Centre for Vision Speech and Signal Processing
Alan Turing Building (BB)
University of Surrey