Rohan Choudhury

prof_pic.jpg

I’m a PhD student in the Robotics Institute at Carnegie Mellon University, advised by Kris Kitani and László Jeni. I’m supported by the NSF GRFP Fellowship, and have spent time during my PhD at Meta FAIR. working with Jing Huang.

I’m broadly interested in making video understanding faster and more efficient. In particular, I work on enabling vision algorithms to continuously perceive the world in high resolution and at 30+ FPS.

Before starting my PhD, I was a software engineer at Nuro, where I worked on trajectory forecasting models for self-driving vehicles. Even before that, I graduated from Caltech with a B.S. in Computer Science, where I worked on multi-agent reinforcement learning with Yisong Yue.

Outside of research, I enjoy running, lifting weights, watching sports and listening to electronic music.

news

Oct 18, 2024 Our work RLT was accepted to NeurIPS 2024 as a spotlight paper!
Jul 14, 2024 Our paper Video Question Answering with Procedural Programs was accepted to ECCV 2024!

selected papers (full list)

  1. rlt_scrat.gif
    Don’t Look Twice: Faster Video Transformers with Run-Length Tokenization
    Rohan Choudhury, Guanglei Zhu, Sihan Liu, and 3 more authors
    NeurIPS, 2024
  2. ECCV
    Video Question Answering with Procedural Programs
    Rohan Choudhury, Koichiro Niinuma, Kris M. Kitani, and 1 more author
    ECCV, 2024
  3. ICCV
    TEMPO: Efficient Multi-View Pose Estimation, Tracking, and Forecasting
    Rohan Choudhury, Kris M. Kitani, and László A. Jeni
    ICCV, 2023