Dilli Bhaskar
M.A.Sc Candidate, Electrical & Computer Engineering, University of British Columbia
I'm a 1st year master's student in Electrical and Computer Engineering at UBC, advised by Prof. Cyrus Neary. Previously, I worked at IIT Madras and IIT Palakkad on Deep RL for robotic autonomy, including UAV pursuit–evasion and robot motion planning, with Prof. Chandrashekar Lakshminarayanan and Prof. Shaikshavali Chitraganti. I completed my Bachelor's in Mechanical Engineering at Thiagarajar College of Engineering (TCE) under the guidance of Prof. S. Saravana Perumaal and Prof. C. Selva Kumar.
I'm interested in developing physics-informed RL algorithms that enable robots to learn robust control policies directly from visual observations. My research focuses on combining world models, representation learning, and physical priors to build autonomous systems that are both data-efficient and physically grounded.
- How can we incorporate physical laws and inductive biases into latent world models learned from visual observations?
- Can physics-structured latent representations improve sample efficiency and generalization in model-based reinforcement learning?
- How can vision and proprioceptive signals be jointly integrated to enable robust control and planning in real-world robotic systems?
