C-Lab: Cognitive systems

Machines that can learn, reason and understand the world.

Man using robot

Current activities

  • Robotics
  • Learning
  • Human computer interaction
  • Man machine interfaces.

Current projects

  • Swiss Nation Science Foundation: SMILE (2016-2019) “Scalable Multimodal sign language
  • Technology for sIgn language Learning and assessment”
  • H2020: Context4ALL (2017-2020) “Personalised Content Creation for the Deaf Community in a Connected Digital Single Market”
  • InnovateUK: AVP (2018-2021): “Autonomous Valet parking
  • Engineering and Physical Sciences Research Council: ExTOL (2018-2021) "End to End Translation of British Sign Language"
  • Four industrial PhDs with: Tesco Labs, Raymarine, McLaren, Jaguar Land Rover.

Previous projects

EU

  • CogViSys
  • COSPAL
  • URUS
  • DIPLECS
  • Dicta-Sign.

EPSRC

  • LILiR
  • Making Sense
  • Learning to Recognise Dynamic Visual Content.

Distinctive approach and achievements

  • Both fundamental and application driven research
  • Weak supervision pervades approaches
  • Human inspired
  • First ever sign language translation work
  • SOTA performance on many gesture and tracking benchmarks
  • Many best paper and thesis awards
  • Consistent record of high quality TPAMI papers over the last three REF periods.

Future focus

  • New project on level four autonomy
  • Self driving cars and boats
  • Strengthening links with the Department of Mechanical Engineering Sciences and Surrey Space Centre to help build critical mass in this area
  • 3D perception and reasoning over scene content
  • Human pose estimation and prediction
  • Sign to Sign Translation System.

Partners

Logos of all the CVSSP partners