Ashok Chandrashekar
 
 
 
Greetings,
 
I am a masters student in the computer science department at Dartmouth College and am associated with the Brain Engineering lab, Neukom Institute. My advisor is Professor Richard Granger.
 
I have a bachelors in computer science from R.V.C.E, Bangalore and have worked on internet routing protocols at Huawei R&D center for a couple of years before landing here in beautiful Hanover.
 
Contact:
Come visit me at: 
116, Sudikoff

email:
<firstname><dot><lastname>@dartmouth.edu

Address:
Ashok Chandrashekar
HinmanBox:6211,
Hanover, NH -03755

Tel : 1-603-646-9797
News:
 
We won the grand prize in the inaugural IBM CELL university challenge!
Here is the official press release and a few quotes from Andrew on eetimes.
 
Publications:
 
Accelerating Brain Circuit Simulations of Object Recognition with a Sony PlayStation 3   -  Felch, A., Moorkanikara-Nageswaran, J., Chandrashekar, A. , Furlong, J., , Dutt, N., Nicolau, A., Veidenbaum, A., Granger, R. H. (2007).  Proceedings of the International Workshop on Innovative Architectures. [pdf]
 
A Brain Derived Vision System Accelerated by FPGAs  - Furlong, J., Felch, A., Moorkanikara-Nageswaran, J., Chandrashekar, A., Dutt, N., Nicolau, A., Veidenbaum, A., Granger, R. H. (2007). Proceedings of ParaFPGA Conference: Parallel Computing with FPGA's. [pdf]
 
draft-holla-ospf-update-graceful-restart - Chandrashekar.A,, Kumar.A,, Gupta.S. (2006 - link)
Resume:  [pdf]
Research Interests:
 
I am in general interested in tasks thats have been traditionally considered AI. I strongly believe that biological mechanisms can provide valuable guidance in solving some of the fundamental problems in AI. In this direction, I am applying brain derived models towards computer vision for object recognition. I am also interested in traditional AI approaches based on numerical and statistical methods.
 
Success on AI tasks often depend on computer architecture advances, and these are exciting times!  The current trend towards increased hardware parallelism, is providing us with wonderful opportunities for approaching tasks which were previously intractable on traditional hardware. Brain derived models in general are highly parallel and scale very well using parallel hardware.