Ashwin Thangali

I am a Ph.D. candidate in the Image and Video Computing Group (IVC) at Boston University. My thesis work relates to exploiting phonological constraints for handshape recognition in sign language video with a goal towards building a person independent large vocabulary sign recognition system. We propose an image alignment based similarity measure for hand location detection and handshape retrieval in sign language video to gain increased robustness to handshape appearance variations.

I have also collaborated on research related to object detection wherein the detector is parameterized to handle foreground class variations, and on person tracking in cluttered scenes (for example in parking lots) where we leverage apriori computed constraints on person motion introduced by occluding objects (such as cars) in the scene.

My advisor is Prof. Stan Sclaroff. My research involves close collaboration with Prof. Carol Neidle's group for their sign language linguistics expertise.


Boston University
Department of Computer Science
111 Cummington St.
Boston, MA 02215