About Me
I build multimodal AI that understands long-form video: what happens, when, who is speaking, and what matters to the viewer.
I am a Senior Researcher in the Multimodal Experiences Lab of the Advanced Technology Group at Dolby Laboratories in San Francisco. I received my PhD in Electrical Engineering from New York University in 2022, advised by Chinmay Hegde. Before that I studied at Iowa State University and BITS Pilani, Goa, where I earned degrees in Electrical Engineering and Physics.
At Dolby, I work on video moment retrieval, long-form video question answering, event-level scene graphs, temporal action localization, active speaker detection, and agentic workflows for personalized video curation.
During my PhD I worked on inverse imaging problems, including compressed sensing, phase retrieval and HDR imaging with deep generative and untrained network priors, with provable guarantees. I also developed algorithms for training adversarially robust networks and studied new attacks on vision transformers.
News
- Sep 2026Semantic Action Graph (SportSAGE) accepted at the IEEE VIS 2026 Workshop on GenAI, Agents, and the Future of VIS. arXiv
- Jul 2026Two papers accepted at ECCV 2026: OnPoint on point-supervised online action localization, and Video2Reaction on predicting audience reactions to video.
- Apr 2026Our work on mitigating hallucinations in vision-language models was accepted to Findings of ACL 2026. Paper
- Jun 2025Presented Moment Sampling in Video LLMs for Long-Form VQA at the CVPR 2025 Workshop on Video LLMs.
- 2022Completed my PhD at NYU and joined Dolby Laboratories as a Senior Researcher.