Sagar Gubbi Venkatesh

Indian Institute of Science Bangalore, Robert Bosch (India)

Papers

Total Citations

H-Index

About

Sagar Gubbi Venkatesh is a researcher at the forefront of human-robot interaction, focusing on how robots can learn from and collaborate with humans in unstructured environments. His work centers on three key challenges: enabling robots to recognize novel objects, understand natural language instructions, and reason spatially. His most cited paper, "One-Shot Object Localization Using Learnt Visual Cues via Siamese Networks" (10 citations), introduces an end-to-end neural network that allows a robot to localize a previously unseen object specified by a visual cue, a critical capability for operation in novel settings. In "Translating Natural Language Instructions to Computer Programs for Robot Manipulation" (9 citations), he proposes a novel approach that converts natural language commands into executable computer programs, moving beyond direct actuator prediction to more robust, interpretable control. His work "Teaching Robots Novel Objects by Pointing at Them" (5 citations) explores intuitive human teaching methods, while "Spatial Reasoning from Natural Language Instructions for Robot Manipulation" (3 citations) tackles the complex task of grounding spatial language in physical scenes. Together, Venkatesh’s research advances the vision of robots that can seamlessly integrate into human environments, learning and adapting through natural communication.