Shailendra Sekhar

Research

Autonomous driving, across the stack

I explore problems all the way from seeing the road to deciding what to do on it. The common thread is that roads are shared with people, and people are hard to predict.

  • Perception
  • Planning
  • Navigation
  • Multi-agent systems
  • Driver behaviour
  • Reinforcement learning
  • Explainable policies

Publications

Google Scholar ↗
  1. ICRA 2026

    Imitation-BT: Automating Behavior Tree Generation by Echoing Reinforcement Learning Agents

    Shailendra Sekhar Bathula, Ramviyas Parasuraman

    Distils deep RL policies (DQN, PPO, A2C) into behavior trees you can read. Retains up to 100% of the expert's reward, often with an order of magnitude fewer nodes than prior BT synthesis.

  2. M.S. Thesis

    Synthetic Instincts: Echoing Reinforcement Learning Agents for Behavior Tree Generation

    Shailendra Sekhar Bathula

    M.S. thesis, University of Georgia, advised by Prof. Ramviyas Parasuraman.