Goku
l Kannan
PhD student at Johns Hopkins University in the VIU Lab, working on robot learning.
exhausted by the clanker
About
My research focuses on advancing robot learning by developing reward models, critics and infrastructures that can enable scalable post-training.
Research
Website
arXiv · coming soon
Archive
Undergrad Research & Projects
Publications
SurgicalGPT: End-to-End Language-Vision GPT for Visual Question Answering in Surgery
MICCAI 2023
Relational Reasoning in VQA Models for Visual Question Answering in Robotic Surgery
ICRA 2023 Workshop · Surgical Robotics
Deep Survival Analysis and Counterfactual Inference Using Balanced Representations
ICASSP 2023
Projects
TANR2L: Transfer of Autonomy using Neurosymbolic Representation-based Reinforcement Learning
Research on transferring autonomy through neurosymbolic representation-based reinforcement learning methods.
Crypto-RL
Developed cryptocurrency trading OpenAI Gym environments for discrete action space with flexible action and state space, with baseline DRL algorithms for benchmarking.
Camoscanner
Built a deep learning pipeline to detect camouflaged soldiers in real-time using an NVIDIA Jetson Nano.
