I lead the neural networks vision group at Vicarious. I completed my PhD with Jim Rehg at the School of Interactive Computing, Georgia Tech.
What excites me is that we know of at least one exceptional vision system: the brain - which theoretically gives Artificial Intelligence researchers a system to mimic. Interestingly, some neuroscientists have shown the visual cortex relying heavily on motion to recognize patterns [1]. This is why I like working with video data. We humans have developed our perception from a temporally continuous stream of information - not from individual images - so why should the machines we make be any different? Over the coming years I would like to build algorithms which can perceive video through machine learning. I would like these algorithms to be flexible enough to even comprehend single images.
Over the years, I have worked on segmentation and occlusion. I am currently intrigued by problems where improvements in combinatoric optimization can help either make problems tractable or reveal more information about video sequences.
Apart from Computer Vision, I have had interludes into Systems research - working on Google's MapReduce with Umar Saif.
Biography: I received my Ph.D. in computer vision at the School of Interactive Computing, Georgia Institute of Technology in 2018. My thesis explored recognition problems in videos where motion and sparse labeling can be used to build a life-long object learning system. Before attending Georgia Tech, I received my Masters degree in CG, Vision & Imaging in late 2010 from UCL. Here, I researched with Gabriel Brostow (aka Gabe) on detecting regions of occlusion in consecutive video frames. I also did a brief stint at The University of Warwick with Nasir Rajpoot, developing registration and dimensionality reduction techniques for cancerous tissue examined under Toponome Imaging System. Previously, I was stationed at LUMS SSE where I worked with Sohaib Khan. In my 3 years stay, I collaborated with biologists at LUMS SSE and MRC NIMR in developing tracking techniques for fluorescence microscopy.