Research
My main research interests lie at learning robust representations, safe policies, and hierarchical skills for long-horizon manipulation and navigation tasks. Following are some questions which motivate my research interests:
- How can we learn robust object-centric representations so that agents can safely perceive and act in the world?
- How to learn hierachical skill policies for effective re-usability and generalization?
- How can we provide safety guarantees for the skills learnt?
Publications
-
Guiding Offline Reinforcement Learning Using Safety Expert, OfflineRL Workshop, NeurIPS 2022
Richa Verma, Kartik Bharadwaj, Harshad Khadilkar, Balaraman Ravindran
-
Continuous Tactical Optimism and Pessimism, Adaptive Learning Agents Workshop, AAMAS 2023
Kartik Bharadwaj, Chandrashekar Lakshminarayanan, Balaraman Ravindran
Awards and Achievements
- 2nd Prize, Third RBCDSAI Conference on Deployable AI, IIT Madras, June 2023 | for presenting my research work on Continuous Tactical Optimism and Pessimism
- 2nd Prize, RBCDSAI Annual Research Showcase, IIT Madras, June 2022 | for present our research work to industry leaders working in ML and Data Science
- Special Achievers Award, VIT Vellore, March, 2018 | Granted by the university for my work in Assailing Falcons
- Among 75 teams, ranked 1st in India and 7th worldwide with 4th in Technical Report & Presentation, SAE Aero design Competition, Florida, USA, April, 2017 | Assailing Falcons
News
- Dec 2017 | Started 6-month internship at Robert Bosch, Bengaluru | Working on medical imaging to detect abnormal RBCs, WBCs, and platelets morphology
- Sep 2020 | Started MSCS at IIT Madras
- June 2023 | Started research internship at UMass Amherst with Prof. Bruno da Silva
|