Ayush Gaud

Indian Institute of Technology Hyderabad, International Institute of Information Technology, Hyderabad

Papers

Total Citations

H-Index

About

Ayush Gaud’s research lies at the intersection of robotics, computer vision, and autonomous systems, with a primary focus on visual servoing—the use of image data to guide robot motion. His major contributions address a critical limitation of classical visual servoing: its reliance on specific object instances. Gaud pioneered instance-invariant frameworks that enable robots to navigate and inspect novel objects without prior training, a breakthrough for real-world applications like autonomous vehicle inspection. His 2019 paper on this topic, with 11 citations, proposes a part-aware approach using micro aerial vehicles (MAVs) to inspect unknown parts of a vehicle, demonstrating robust performance across varying object shapes. In his 2018 work (9 citations), Gaud tackled the challenging problem of visual servoing toward tumbling objects in space, introducing methods that bypass explicit 3D reconstruction—a key step for uncooperative space debris handling. His 2017 paper (5 citations) further advanced the field by formulating pose induction for servoing to entirely novel object instances, eliminating the need for instance-specific models. Gaud’s work is notable for its practical impact, bridging the gap between controlled lab settings and dynamic, unstructured environments, and has been cited by researchers in aerial robotics, space exploration, and industrial automation.