In addition to general lab space our Centre hosts a number of unique research facilities.
Visual Media Laboratory
The Centre has created the Visual Media Laboratory - a state-of-the-art video and audio capture studio that supports research in real-time video and audio processing and visualisation. The main feature of the lab is a network of eight high-definition, broadcast-quality cameras each with an associated grabbing and storage system.
We also have more portable HD cameras which can be networked to either capture to tape or to our mobile capture system. This is supported by an extensive lighting rig. The cameras and display devices are linked to a reconfigurable SDI network, which also supports audio capture and playback.
The Laboratory is also equipped with a 4sensor Coda MotionTMcapture system and two face capture systems: A commercial 3D static capture system; and an in-house dynamic 3D capture system.
The Surrey Sound Sphere is the centrepiece of the cutting edge audio and acoustics research taking place at the University of Surrey.
The system consists of 64 Genelec loudspeakers, together with a Mac Pro, sound cards to interface with the loudspeakers, and preamplifiers to simultaneously record 48 channels of audio from a microphone array. There's over a kilometre of cables to connect everything together!
The sphere has been used for experiments on personal sound zones (where you hear a completely different piece of audio depending on where you stand), testing human sound localisation (working out what affects your ability to sense where sound is coming from), and is currently in use as a production environment for object-based 3D audio.
Also housed within our Centre is the Robot Lab: A reconfigurable space to support research into autonomous systems, linking closely with the Centre's expertise into computer vision, AI and machine learning.
It includes a safe area for indoor autonomous aerial vehicles (UAVs) to support research into collaborative mapping based on visual simultaneous localisation and mapping (SLAM). Additionally, a variety of robotic platforms support areas of research such as autonomous navigation, perception-action learning and human computer interaction.
To support the laboratory hardware, we have an array of servers running Ubuntu Linux totalling over 700CPUs, several state of the art GPUs and around 4TB of RAM. In order to store the data for processing, the Centre has in the region of 0.5PB of networked storage mirrored across two buildings for security - this is augmented by a recently purchased, university wide storage solution to double the current provision and offer higher speeds of access.