PeopleofSuperintelligence

Researcher

David Ross

David Ross is a computer vision researcher whose work spans latent-factor models, human action recognition and video generation. He earned a University of Toronto Ph.D. in machine learning and computer vision in 2008, then helped develop AVA, a dataset that locates atomic actions in video. In 2021, he coauthored AIST++ for music-conditioned dance generation; his 2023 VideoPoet work used a large language model to generate video, and the paper won ICML's 2024 Best Paper award. He coauthored MALT Diffusion for long video generation in 2025, and he co-leads Google DeepMind's VIVID research group on video understanding and generation.

Badges

Timeline

  1. 2006

    Learning parts-based models

    With Richard Zemel, he publishes Learning Parts-Based Representations of Data in JMLR in 2006. The probabilistic model learns parts and their possible states, and applies them to images of faces and movie ratings.

  2. 2008

    A Toronto computer vision doctorate

    He earns a University of Toronto Ph.D. in machine learning and computer vision in 2008 with a thesis on probabilistic visual motion.

  3. 2017

    AVA labels human actions

    In October 2017, he introduces AVA with Google Research colleagues as a dataset of 80 atomic visual actions localized in space and time. Its annotations assign multiple action labels to people in extended video sequences.

  4. 2021

    AI Choreographer and AIST++

    In 2021, he coauthors AI Choreographer, a full-attention crossmodal Transformer and the AIST++ 3D dance-and-music dataset. The model generates long sequences of dance motion conditioned on music.

  5. 2023

    VideoPoet for video generation

    In 2023, he introduces VideoPoet with Dan Kondratyuk, a large language model for text-to-video, image-to-video, stylization, inpainting, outpainting and video-to-audio.

  6. 2024

    An ICML Best Paper

    He coauthors VideoPoet, which wins ICML's 2024 Best Paper award. Its decoder-only language model processes images, videos, text and audio for zero-shot video generation.

  7. 2025

    MALT Diffusion for long videos

    In 2025, he coauthors MALT Diffusion, which uses a compact memory across short segments to generate video over longer temporal contexts. The work is an oral presentation at the CVPR 2025 Workshop on AI for Content Creation.

  8. As of October 2026

    Co-leading VIVID at Google DeepMind

    He co-leads the VIVID research group at Google DeepMind, whose goal is to advance video understanding and generation. In June 2025, he is a Research Manager at Google DeepMind.

Career

  1. Since 2025Google DeepMindResearch Manager and VIVID Co-lead

Education

  1. 2008University of TorontoPhD

From the biography.

  1. The median h-index is 48 at universities and 14 at frontier AI labsIncludes David Ross. 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.
  2. 16% of AI career moves in the 2020s went from a university to a companyIncludes David Ross. We estimate that 16% of AI career moves in the 2020s went from a university to a company; 29% of moves starting at a university went to a company, and 76% ended the university role.
  3. The Massachusetts Institute of Technology trained or employed 44 of the 187 AI major-prize winners, more than any other universityIncludes David Ross. The Massachusetts Institute of Technology trained or employed 44 of the 187 AI major-prize winners, more than any other university. We compare doctorate and 5+ year employment links, career timing and current workplaces for 187 major-prize and 1,111 best-paper winners.
  4. Google had the most people shaping AI every year from 2014 to 2025Includes Google. Our estimates of which organizations employed the most people shaping AI each year from 1955 to 2025, tracing the shift from university labs to companies.

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 David Ross yet. Send us a free-licensed photo.

  1. MCVQ and MCFA cs.toronto.edu. Checked 21 October 2026.
  2. David Ross ross-da.github.io. Checked 21 October 2026.
  3. Announcing AVA: A Finely Labeled Video Dataset for Human Action Understanding research.google. Checked 21 October 2026.
  4. AI Choreographer: Music Conditioned 3D Dance Generation with AIST++ research.google. Checked 21 October 2026.
  5. VideoPoet: A large language model for zero-shot video generation research.google. Checked 21 October 2026.
  6. ICML 2024 Awards icml.cc. Checked 21 October 2026.
  7. MALT Diffusion: Memory-Augmented Latent Transformers for Any-Length Video Generation arxiv.org. Checked 21 October 2026.
  8. Sixth International Workshop on Large Scale Holistic Video Understanding holistic-video-understanding.github.io. Checked 21 October 2026.