|
Abstract: Despite their seemingly impressive performance at image recognition and other perceptual tasks, deep convolutional neural networks are prone to be easily fooled, sensitive to adversarial…
|
|
Abstract: Flexible function is essential for the brain to cope with varying environments, changing quality of sensory information as well as context dependency. This requires organized communication…
|
|
Abstract: As Moore’s
Law comes to a close, new innovative approaches to microelectronics research
are important to achieve much needed capabilities improvements in computing for
both…
|
|
Prof. Cawenberghs presents neuromorphic cognitive computing systems-on-chip implemented in custom silicon compute-in-memory neural and memristive synaptic crossbar array architectures that combine…
|
|
Discussion of analog CMOS computation and parallels to biological neurons. Prof. Hasler points out that the exponential voltage controls found in MOS devices can be used to more accurately model…
|
|
Luis El Srouji leads a discussion on Vector Symbolic Architectures (VSA). VSAs are a mathematical framework in which basis vectors encode some attribute (color, weight, etc) while some binding…
|
|
Professor Olshausen discusses the neural circuits thought to underly signal compression in the retina, and theories of attractor dynamics downstream in cortex that may underly perceptual inference.
|
|
Professor Chris Eliasmith gave a guest presentation on some mechanisms of encoding information within spiking neural networks. Spatial Semantic Pointers are an extension of the semantic pointer…
|
|
Algorithmic Frameworks for Neuromorphic Computing: Neural Engineering Framework (NEF) Luis El Srouji gives an overview and explanation of the Neural Engineering Framework which can calculate weight…
|
|
Professor Randy O'Reilly presented on Predictive Error-Driven Learning in the Brain; this method approximates backpropagation in rate-approximated neural networks and will be soon extended to…
|
|
A review of basic neurobiology followed by a discussion on the capabilities of IBM's TrueNorth and Intel's Loihi neuromorphic chips.
|
|
First club meeting; a brief overview of neuromorphic computing in comparison to both biological brains and Von Neumann computers. A short discussion of ABR's Nengo library and Intel's Loihi…
|