Sign Finder
This project seeks to develop a computer vision-based system that allows a visually impaired traveler to find and read informational signs, such as signs labeling office doors, streets, restrooms and Exit signs.
Link to open source code
This project seeks to develop a computer vision-based system that allows a visually impaired traveler to find and read informational signs, such as signs labeling office doors, streets, restrooms and Exit signs.
Link to open source code
These are tutorials and reference materials I have written on various topics in probability and geometry over the years.
This project study the neural mechanisms of texture and motion based segmentation as basic inputs to object processing. This project used frequency-domain non-linear analysis of high-density EEG recordings in human infants. We also used fMRI-informed EEG source-imaging combined with psychophysics to study texture segmentation processing and the role of attention in normal adults and in adults with amblyopia.
This project studies the integration of local estimates of motion direction and orientation into global patterns. Integration mechanisms were studied in human infants, normal adults and in adults with amblyopia.