home | members | publications | directions | research | presentations | teaching | vacancies | personal
Particular current focusses of the Computational Perception Group are: deep probabilistic learning, human-like learning, continual learning, k-shot learning, active learning, reinforcement learning, concept learning, Bayesian neural networks, Gaussian processes, spatio-temporal modelling, approximate inference, scalable and distributed inference, Monte Carlo methods, variational methods, expectation propagation, and Bayesian optimisation. We are also interested in the connection between machine learning and computation in the brain.
Please see the Publications section for an indication of relevant recent activity in these areas.
Previous and ongoing research includes the following:
Machine Perception
Statistical models for audio and video
Theoretical understanding of learning algorithms as probabilistic inference
Machine Vision
Learning invariances from natural images for object recognition
Statistical models for images
Machine Hearing
Synthesis of audio textures for computer games and artificial environments
Source separation
Neuroscience
Auditory processing as probabilistic inference
Neural implementations of approximate inference
Approximate inference for time-series
Circular statistics and time-series
Signal Processing
Unifying signal processing and machine learning
Removing signal distortions using machine learning & signal processing
If you want to find out more about the core technical material that is relevant to the Group's reseach, please see the Group's recommended reading list.