Kwabena Boahnen and his work

Read an interesting article in the Discover magazine about the work of Kwabena Boahnen and his team from Stanford.

http://discovermagazine.com/2009/oct/06-brain-like-chip-may-solve-computers-big-problem-energy/article_view?b_start:int=0&-C=

I think the approach taken by them is pretty interesting and might result in new kinds of circuits inspired from the principles of brain. His approach might have interesting application in sensor processing or in other applications. But when coming to proper neuroscience applications there are one main shortcoming in their approach. Brain-science research is one of the most fast moving fields with most ideas having a very short-life period. Thus what is currently needed is some cheap, massive programmable multiprocessor platform like the work done by Steve Furber Group's SPINNAKER. This has much value to really test drastic ideas rather than a fixed hardware platform similar to neurogrid.

Book review: Synaptic Self by Joseph Ledoux

The past three weeks I am going over an interesting book called "Synaptic Self - Joseph LeDoux". Broadly, books about neuroscience fall into two different categories, layman books and formal textbooks. Layman books discusses about day-to-day aspects of brain mechanisms like sleep, addiction, memory etc. The second category is the text books that are extremely technical in nature (Kandel Principles etc). Given my urge to read and understand about brain as much as possible, I encountered Joseph LeBoux's book that targets Scientific American level readers. The concepts are all extremely well written and the language/style is simply superb. The book mixes both historic references along with latest discoveries and establishes a perfect link between them. Even though my interest lies in computational neuroscience aspects of the brain, this book helped me in getting a better picture about other brain areas that I have not often read or worked on like amygdala, hippocampus, cerebellum, and so on. This book is a must read for any aspiring neuro-scientist or neuromorphic engineers.