PeopleofSuperintelligence

Researcher

Pritam Sarkar

Also known as প্রীতম সরকার

Pritam Sarkar is a multimodal AI researcher who designs algorithms for video, image, audio and language models that learn with minimal human supervision. He defended a master's thesis on self-supervised ECG representation learning at Queen's University in 2020, then completed a PhD there in 2025 under Ali Etemad, with internships at Borealis AI and Google along the way. His papers include CrissCross on audio-visual learning at AAAI 2023, a NeurIPS 2023 spotlight on video self-supervision under distribution shifts, and the RRPO method for aligning video language models at NeurIPS 2025. In 2026, the Canadian Image Processing and Pattern Recognition Society gave him its John Barron Doctoral Dissertation Award for the PhD thesis. He is a Distinguished Postdoctoral Fellow at the Vector Institute and a postdoctoral research fellow at the University of British Columbia, where he works with Leonid Sigal.

Badges

Timeline

  1. 2018–2020

    A master's on ECG signals

    He joins Queen's University in September 2018 for a master's degree and defends his MASc thesis in April 2020. Supervised by Ali Etemad in Electrical and Computer Engineering, the thesis develops a self-supervised ECG representation learning framework for affective computing. The related paper with Etemad sets new state-of-the-art results in classifying arousal, valence, affective states, and stress on four datasets, and is accepted in IEEE Transactions on Affective Computing.

  2. 2020–2025

    PhD at Queen's University

    In 2020, he begins a PhD in Electrical and Computer Engineering at Queen's University under Ali Etemad. His thesis, Multimodal Learning from Videos, spans self-supervised pre-training, post-training alignment with human preferences, and new multimodal video benchmarks; he defends it in September 2025.

  3. 2022–2023

    Borealis AI and Google internships

    He spends fall 2022 as a machine learning research intern at Borealis AI in Toronto and fall 2023 as a student researcher at Google in Sunnyvale.

  4. 2023

    CrissCross and a NeurIPS spotlight

    At AAAI 2023, he and Ali Etemad present CrissCross, a self-supervised framework that learns audio-visual representations by relaxing the temporal synchronicity between the two modalities. With Ahmad Beirami and Etemad, he also studies how six video self-supervised methods behave under six kinds of natural distribution shift, in a paper accepted as a NeurIPS 2023 spotlight.

  5. 2023

    An IEEE Kingston research prize

    He wins first prize in the Research Excellence Award (PhD) of the IEEE Kingston Section in 2023, for his contributions to self-supervised learning.

  6. 2025

    RRPO for video language models

    In 2025, he and Etemad introduce Refined Regularized Preference Optimization (RRPO), a self-alignment method that lets large video language models learn from their own errors. It uses sub-sequence-level refined rewards and token-wise KL regularization to address limits of Direct Preference Optimization, and the paper is accepted at NeurIPS 2025.

  7. 2026

    A Barron dissertation award

    In April 2026, he receives the John Barron Doctoral Dissertation Award from the Canadian Image Processing and Pattern Recognition Society, given annually to the top PhD thesis in the areas covered by the Conference on Computer and Robot Vision. The award recognizes his Queen's University thesis in the computer vision category.

  8. As of October 2026

    Postdoc at UBC and Vector

    He joins the University of British Columbia and the Vector Institute as a postdoc in December 2025, working with Leonid Sigal in the Computer Vision Group, and holds a Vector Distinguished Postdoctoral Fellowship. His research covers multimodal AI across video, image, audio, and language, and he also serves as a research advisor to startups and companies working on generative AI and computer vision.

Career

  1. Since 2025University of British ColumbiaPostdoctoral research fellow
  2. Since 2025Vector InstituteDistinguished Postdoctoral Fellow
  3. 2023GoogleStudent researcher
  4. 2022Borealis AIMachine learning research intern

Education

  1. 2025Queen's UniversityPhD
  2. 2020Queen's UniversityMASc

From the biography.

  1. Of 943 people shaping AI who worked at Google, 446 work there nowIncludes Pritam Sarkar. Our estimates of where the 943 people shaping AI who worked at Google and Google DeepMind work now: 446 at Google and 452 across 278 other organizations.
  2. The median h-index is 48 at universities and 14 at frontier AI labsIncludes University of British Columbia. The median h-index is 48 at universities and 14 at frontier AI labs. We compare publishing careers, recent citations, dual appointments and job moves across 2,420 university researchers and 308 frontier-lab researchers.
  3. Nonprofit roles are 2.3 times as common in AI safetyIncludes Vector Institute. We estimate where people focused on safety, alignment and evaluation work. Nonprofit roles appear for 25.1% of safety people with known affiliations, compared with 11.0% in other areas.
  4. Universities (31%) and startups and other companies (25%) employ the largest shares of 2020–2026 entrantsIncludes Vector Institute. Among 819 people whose first recorded role started in 2020–2026 and who have a known current organization, 60% work in commercial sectors (25% at startups and other companies, 23% at frontier labs and 13% at big tech, with 9 in two), and 31% work at universities, compared with 66% for earlier entrants.

Colleagues

People named in this biography, and people with dated positions at the same organization in the same years.

Sources and credits

Photo: We have no published photo of Pritam Sarkar yet. Send us a free-licensed photo.

  1. Pritam Sarkar pritamsarkar.com. Checked 22 October 2026.
  2. Self-supervised ECG Representation Learning for Affective Computing queensu.scholaris.ca. Checked 22 October 2026.
  3. [2002.03898] Self-supervised ECG Representation Learning for Emotion Recognition arxiv.org. Checked 22 October 2026.
  4. Multimodal Learning from Videos: Self-supervised Pre-training, Post-training Alignment, and Benchmarks queensu.scholaris.ca. Checked 22 October 2026.
  5. Pritam Sarkar - Vector Institute for Artificial Intelligence vectorinstitute.ai. Checked 22 October 2026.
  6. Self-Supervised Audio-Visual Representation Learning with Relaxed Cross-Modal Synchronicity | Proceedings of the AAAI Conference on Artificial Intelligence ojs.aaai.org. Checked 22 October 2026.
  7. [2306.02014] Uncovering the Hidden Dynamics of Video Self-supervised Learning under Distribution Shifts arxiv.org. Checked 22 October 2026.
  8. [2504.12083] Self-alignment of Large Video Language Models with Refined Regularized Preference Optimization arxiv.org. Checked 22 October 2026.
  9. CIPPRS cipprs.org. Checked 22 October 2026.