Siddharth Ancha is a postdoctoral research in the Computer Science and Artificial Intelligence Lab (CSAIL) at the Massachusetts Institute of Technology working at the intersection of Robotics, Machine Learning and Computer Vision. His research focuses on developing algorithms for enabling robots to actively, robustly, and adaptively perceive and act in complex environments. He received his Ph.D. from the Machine Learning Department at Carnegie Mellon University. He has published at venues such as RSS, CoRL, IROS, NeurIPS, CVPR, ECCV and 3DV. He is the recipient of the IROS Outstanding Reviewer Award in 2022.
Robots face many challenges in sensing and perception while navigating in the real world. Robots, like people, must actively think about where to look and how to acquire information. Robots must be robust to unseen environments and provide safety and performance guarantees. Instead of relying on large, expensive, human-labeled datasets, they must constantly adapt and update their models as they navigate through dynamic, non-stationary environments where the data distribution changes constantly. My research argues that perception for embodied cyber-physical systems requires more than computer vision. I combine modern deep learning with classical robotics techniques like motion planning, Bayesian inference, uncertainty estimation, information theory and decision-making under uncertainty to enable active, robust and adaptive perception while advocating for a tighter integration between sensing and control in robotics and cyber-physical systems.