You are here

Workshop on Cognition and Control

Some results and challenges in applying active sequential hypothesis testing to a visual search problem
Rajesh Sundaresan
Rajesh Sundaresan received his B.Tech. degree in electronics and communication from the Indian Institute of Technology, Madras, the M.A. and Ph.D. degrees in electrical engineering from Princeton University, NJ, in 1996 and 1999, respectively. From 1999 to 2005, he worked at Qualcomm Inc., Campbell, CA, on the design of communication algorithms for WCDMA and HSDPA modems. Since 2005 he has been an Assistant Professor in the Department of Electrical Communication Engineering at the Indian Institute of Science, Bangalore. His interests are in the areas of wireless communication and information theory. Currently on sabbatical leave at:
Coordinated Science Laboratory, University of Illinois at Urbana-Champaign
I will describe a visual search experiment conducted by Sripati and Olson (2010) where the objective was to identify, as quickly as possible, an oddball image embedded among multiple distractor images. Remarkably, the reaction times for identifying the oddball (in humans) were highly correlated with an ad hoc neuronal metric obtained from measured neuronal responses to component images (in macaques). I will suggest an alternative neuronal metric that is based on an active sequential hypothesis testing model of visual search, and on a classical result of Chernoff on sequential design of experiments (1959). The correlation between the reaction times and the alternative neuronal metric continues to be high, and has the advantage of being firmly grounded in decision theory. I will then discuss some predictions made by our model, and additional experiments that highlight the challenges involved in the modeling. The talk is based on ongoing work with Nidhin Koshy Vaidhiyan and Arun Sripati.
February 22nd
4:30 pm
Reitz Union 346

Theme by Danetsoft and Danang Probo Sayekti inspired by Maksimer