Srivatsan Srinivasan

Papers

Total Citations

H-Index

About

Srivatsan Srinivasan is a researcher at the intersection of machine learning, robotics, and autonomous systems. His work spans Bayesian deep learning, socially assistive robotics, and off-road vehicle autonomy. In his seminal paper "Output-Constrained Bayesian Neural Networks" (2019, 10 citations), Srinivasan introduced a novel framework that encodes functional prior knowledge directly into BNNs, enabling more reliable predictions in safety-critical applications. This work addresses a fundamental limitation of traditional BNNs, where priors are defined in parameter space rather than function space. In robotics, he developed "Misty," a development platform for socially assistive robots (2019, 6 citations), contributing to the growing field of human-robot interaction. More recently, his 2022 paper on deep reinforcement learning for pure-pursuit path-tracking control of skid-steered vehicles (6 citations) tackles the complex challenges of off-road autonomy, where traditional geometric controllers fall short. Srinivasan's research demonstrates a rare ability to bridge theoretical advances in Bayesian inference with practical robotic systems, making his work valuable for both machine learning theorists and robotics practitioners developing autonomous systems for unstructured environments.