Ritik Raina
Ph.D. Candidate, Cognitive Science · Stony Brook University Advised by Gregory Zelinsky Research Scientist Intern · Tavus
Now Building multimodal generative models that are aligned with human visual perception.
Intern Research Scientist Intern at Tavus: video diffusion for real-time conversational avatars.
Latest Eyes of the beholder · CCN 2026 · MetamerGen · ICLR 2026 Oral
About
I am a fourth-year Cognitive Science Ph.D. candidate at Stony Brook University working with Dr. Gregory Zelinsky.
My research interests are in the intersection of multimodal generative modelling and visual perception, focusing on building image and video generative models that are grounded in human visual behavior.
Since May 2026 I have been a Research Scientist Intern at Tavus, where I work on autoregressive video diffusion and distillation for real-time conversational avatars.
I am always interested in research collaborations. Feel free to schedule a 30-minute meeting if you would like to collaborate or talk through research ideas.
News 18
Scroll for more- Aug 2026Two papers accepted at CCN 2026, one selected for a talkNew
- May 2026Started as a Research Scientist Intern at Tavus
- Jan 2026Paper accepted at ICLR 2026 as an oral (top 1%)Oral
- Aug 2025Short paper accepted at CCN 2025
- Feb 2025Talk accepted at VSS 2025
- Nov 2024Poster at OPAM 2024Best Poster
- Aug 2024Short paper accepted at CCN 2024
- Feb 2024Talk accepted at VSS 2024
- Sep 2023Paper accepted at NeurIPS 2023
- Aug 2023Started my Ph.D. at Stony Brook University
- Mar 2023Article accepted at Journal of Vision
- Dec 2022Paper accepted at NeurIPS 2022 SyntheticData4ML Workshop
- Jun 2022Joined the de Sa Lab as a Pre-Doctoral Researcher
- Jun 2022Graduated from UC San Diego with a B.S. in Cognitive Science, minor in Computer Science
- Jan 2022Started as a Student Researcher at Intel
- Dec 2021Paper accepted at NeurIPS 2021 SVRHM Workshop
- Jan 2021Joined the de Sa Lab as a Research Assistant
- Jan 2021Started as an Undergraduate ML Researcher at IBM Research
Publications
-
VJEPAvatar: audio-grounded video pretraining for talking headsComing soon
No matches.
Full list on Google Scholar.
