Sarthak J. Shetty

Papers

Total Citations

H-Index

About

Sarthak J. Shetty is a researcher at the intersection of computer vision, robotics, and intelligent automation. His work spans three key areas: vision-based quality inspection, autonomous drone systems for search and rescue, and robotic manipulation with tools. In his 2019 paper on fracture detection in manufactured components—cited 12 times—Shetty developed a neural network-driven vision system that automates quality control in production lines, addressing the growing need for reliable, automated inspection in Industry 4.0. His 2021 work on drone-based survivor detection, with 10 citations, proposes practical strategies for deploying unmanned aerial vehicles in disaster response scenarios. Most notably, Shetty introduced ToolFlowNet (2022, 7 citations), a novel framework that learns robotic tool manipulation directly from point clouds by predicting tool flow—a significant advance over traditional imitation learning methods that rely on visual or proprioceptive inputs. This work addresses the fundamental challenge of policy learning from 3D geometric data, enabling robots to use tools more effectively in unstructured environments. Shetty’s contributions demonstrate a consistent focus on bridging perception and action, with applications ranging from manufacturing to humanitarian robotics.