Computational Perception Group

home | members | research | publications | presentations | teaching | vacancies | personal

The research carried out in the Computational Perception Group lies at the interface between three fields:

  1. computer perception which builds systems for processing and understanding images, sounds and videos
  2. neuroscience which is the scientific study of the nervous system
  3. machine learning which provides a theoretical framework for learning and making inferences from data
The goal of the research is to develop systems that solve important problems, drawing inspiration from the brain. For example, determining how many sound sources there are in an acoustic scene and what the individual contributions from each source are. There are medical and engineering applications of this work, such as in intelligent hearing aids and cochlear implants for the deaf. Importantly, the behaviour of these algorithms can also be compared to neural processing in the brain in order to better understand what the brain is doing.

We collaborate very closely with both experimental neuroscience groups, who research the neural basis of hearing, hearing aids and cochlear implants, and also industrial partners such as Google.

opening image


23 / 05 / 16 A new draft paper on a unifying framework for sparse GP approximation using Power-EP
24 / 04 / 16 Deep Gaussian Processes showing state-of-the-art results on regression accepted to ICML2016
24 / 04 / 16 Black-box α-divergence Minimization accepted to ICML2016
04 / 04 / 16 Interviewed about AI by BBC 5 Live's science programme The Naked Scientist
01 / 02 / 16 Paper accepted by AISTATS and paper accepted by Hearing Research
25 / 11 / 15 Five papers accepted to NIPS2015 workshops, including two oral presentations and one best paper award
04 / 09 / 15 Three papers accepted to NIPS2015, two with spotlight presentations
06 / 06 / 15 Thang Bui awarded a Google European Doctoral Fellowship
19 / 05 / 15 Richard Turner awarded a Cambridge Students' Union Teaching Award for Lecturing
18 / 03 / 15 Richard Turner awarded an EPSRC Research Grant
08 / 12 / 14 The Group's research features on the BBC 5 Live's science programme, The Naked Scientist
20 / 09 / 14 New journal paper "Time-frequency analysis as probabilistic inference" accepted for publication
09 / 09 / 14 Tree structured Gaussian Process Approximations paper accepted to NIPS2014 with spotlight presentation
28 / 10 / 13 The Group's research features on the BBC World Service's technology programme, Click
07 / 05 / 13 Richard Turner awarded an EPSRC First Grant
22 / 02 / 13 Richard Turner awarded a Google Faculty Award
This site is powered by FoswikiCopyright © by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding Foswiki? Send feedback