Rohan graduated with a B. Tech and M. Tech degree in Electronics and Electrical Communication from IIT Kharagpur in 2014. His masters thesis is based in interesting information theoretic based sparse optimizations for vision. He did his internship in SINAPSE prior to joining as a PhD student. His work in SINAPSE started with invariant object recognition and orientation estimation using Convolutional Networks (CNN).
Such a system is intended for use with a prosthetic arm to aid in grasp planning.
His current interests and endeavours lie in :
Probability generalization theory.
Using Self Organizing Maps (SOM) for min-count-spike encoding in on-chip Extreme Learning Machines for Fast object recognition (in collaboration with NTU)
Apart from his above main interests, he is also actively involved in projects including object following and tracking with the DVS, building dark navigation algorithms, and numerous others to quench his insatiable thirst for knowledge and discoveries.