Manmohan K. Chandraker
Asst Professor, Computer Science and Engineering
Computer vision, 3D reconstruction, scene understanding, autonomous driving, vision for robotics, graphics-based vision.
Chandraker’s research in computer vision develops theoretical frameworks and practical systems for recovering three-dimensional (3D) scene information from 2D images, with interdisciplinary links to computer graphics, visual perception, machine learning and robotics. His work characterizing shape recovery under complex reflectance and illumination establishes powerful generalizations of 3D reconstruction methods such as optical flow, multiview stereo and photometric stereo to handle real-world materials and global light transport, while motivating novel applications in augmented reality and light field imaging. Other work includes simultaneous localization and mapping for robotic navigation, dense reconstruction by incorporating semantic inputs such as object detection, as well as practical global optimization frameworks for non-convex problems in multiview geometric reconstruction, shape from exemplars and non-rigid structure from motion. His research has been published in premier venues and been recognized with the Best Paper Award at CVPR 2014, a PAMI Special Issue on Best Papers from CVPR 2011, the UCSD CSE Dissertation Award for Best Thesis in 2009 and the Marr Prize honorable mention for Best Paper at ICCV 2007.
His work on inexpensive solutions for autonomous driving has established a long-term research direction at NEC and collaborations with leading automobile manufacturers. He led a team that developed highly accurate single-camera systems for real-time visual odometry, 3D object localization, scene understanding and recognition with applications such as collision avoidance, navigation and path planning for advanced driver assistance systems. This research has been published in leading venues and recognized with several patents and awards, including NEC Spot Recognition Awards in 2012 and 2013 and Business Contribution Award in 2015.
Manmohan Chandraker received a B.Tech. in Electrical Engineering at the Indian Institute of Technology, Bombay and a Ph.D. in Computer Science at the University of California, San Diego. Following a postdoctoral scholarship at the University of California, Berkeley, he worked at NEC Labs America in Cupertino as a senior researcher. His research has been recognized with best paper awards in leading academic venues and business contribution awards in industry.