The following projects are being pursued by laboratory members and collaborators, categorized by broad research area.
New in 2015-2016.
- Natural Language Processing
- Deep learning
- Computer graphics based on models of human visual perception
Existing and Past Projects:
1. Computational Neuroscience Modelling
1A. Networks Neuroscience and Neuronal Connectomes
- Stochastic pooling networks: Is this kind of network topology, where redundancy and noise are jointly exploited, to be found in the brain and specifically in the cortex?
1B. Stochastic resonance and intrinsic neuronal noise
Does stochastic resonance occur in-vivo in neural systems?
What is required to build bridges between the idea of "noise benefits" and contemporary neuroscience modelling?
At what "level of description" does stochastic resonance relate to neuronal "computation?
How does short term plasticity, due to stochastic neurotransmitter release in synaptic communication between neurons, affect neural coding?
1C. Cortical neuronal oscillations
- What neuronal mechanisms are required to exist for existing hypotheses regarding communication within the brain to be plausible?
- Can beta oscillations be reproduced in simulated networks of the piriform (primary olfactory) cortex?
1D. Auditory nerve
Is suprathreshold stochastic resonance exploited in the transduction of sounds into action potentials in primary afferent auditory neurons?
Can the high levels of spontaneous action potentials in primary afferent auditory neurons be explained in terms of suprathreshold stochastic resonance?
2. Computational theory and modeling of visual surface perception
- A primary goal of psychology and neuroscience is to understand how sensory and perceptual systems represent information about the outside world. Determining the mathematical form of the representations underlying such systems is known as the representation problem. The central aim of this research project is to describe the mathematical form of the brain representation underlying the visual perception of surface properties, such as lightness, transparency and gloss.
3. Information Theory and Signal Processing with Neuroscience Applications
3A. Biological Signal Processing
- Optimal electrical patterns for cochlear implants
- Automatic classification of sleep stages from EEG data
- Removal of EMG contamination from EEG data
3B. Lossy compression in neural coding
3C. Critical evaluation of the utility of Shannon information theory in neuroscience
3D. Optimal stimulus-response curves and stimulus distributions
4. Stochastic Modelling for Engineering Applications
- FPGA implementation of stochastic neuron models
- Biologically-inspired analog-to-digital converters and stochastic quantization theory
- Optimization of sensor networks
- The two envelope problem with randomized switching
- Communication using a single photon source