PeopleofSuperintelligence

Researcher

Max Bain

Max Bain is a computer vision researcher whose work helps AI systems understand video, speech and language. He coauthored methods for identifying individual chimpanzees in video and datasets for searching films and videos with text. His Oxford research used audio and video to recognise wild chimpanzee behaviour; AutoAD generated movie audio descriptions, and WhisperX added word-level timestamps to long-form speech. After coauthoring multimodal models and a visual-understanding benchmark at Reka, he is a Research Scientist at Google DeepMind. He coauthored the Gemini 2.5 report in 2025, improved WhisperX in 2026 and coauthored a chimpanzee video dataset in 2026.

Works
30
Citations
783
h-index
11
Badges

Publication numbers from OpenAlex, as of 22 October 2026.

Timeline

  1. 2019

    Chimpanzee recognition from video

    In 2019, he coauthors Count, Crop and Recognise, a multi-stage method for identifying chimpanzee individuals in video frames. The paper introduces a dataset for chimpanzee recognition in the wild.

  2. 2020

    The Condensed Movies dataset

    In 2020, he coauthors Condensed Movies with Arsha Nagrani, Andrew Brown and Andrew Zisserman. The dataset uses key scenes from over 3,000 films for story-based retrieval, and the paper appears as an oral presentation at ACCV 2020.

  3. 2021

    Frozen in Time and WebVid-2M

    With Arsha Nagrani, Gül Varol and Andrew Zisserman, he coauthors Frozen in Time in 2021, a joint video and image encoder for text-to-video retrieval. The paper introduces WebVid-2M, a pretraining dataset of more than two million videos with weak captions scraped from the internet.

  4. November 2021

    Wild chimpanzee behaviour recognition

    As a DPhil student in Oxford's Visual Geometry Group supervised by Andrew Zisserman, he helps develop AI models that recognise wild chimpanzee behaviour from audio and video; the study appears in Science Advances in November 2021. The models use footage from Bossou in Guinea and Cantanhez National Park in Guinea-Bissau to detect nut-cracking, eating and buttress drumming.

  5. March 2023

    WhisperX for long-form speech

    With Jaesung Huh, Tengda Han and Andrew Zisserman, he introduces WhisperX in March 2023, a speech recognition system that uses voice activity detection and forced phoneme alignment to add word-level timestamps. Accepted to INTERSPEECH 2023, it ranks first in the EGO4D Audio-Visual Automatic Speech Recognition Challenge, and the team makes the baseline code and models available online.

  6. 2023

    AutoAD and an Oxford DPhil

    In 2023, he coauthors AutoAD, a model that ingests movies and writes audio descriptions as text; the paper is a CVPR 2023 Highlight. His 2023 Oxford DPhil thesis, Understanding Video Through the Lens of Language, is supervised by Andrew Zisserman and explores audio description for visually impaired audiences.

  7. 2024

    Multimodal work at Reka

    At Reka, where he serves as a Member of Technical Staff, he coauthors the April 2024 report on Core, Flash and Edge. The models are trained from scratch to process text, images, video and audio; in May 2024, he coauthors Vibe-Eval, a benchmark with 269 visual-understanding prompts for multimodal chat models.

  8. As of October 2026

    Research Scientist at Google DeepMind

    In July 2025, he is among the authors of the Gemini 2.5 technical report. In August 2026, he and Carlos Bain publish context-aware interleaved batching for WhisperX, which reduces word error rate while maintaining high throughput; in October 2026, he coauthors BossouChimpanzee, a video dataset of wild chimpanzees at Bossou, Guinea, spanning 1988 to 2018. As of October 2026, he is a Research Scientist at Google DeepMind.

Career

  1. Since 2025Google DeepMindResearch scientist
  2. Until 2024Reka AIMember of technical staff

Education

  1. From 2021University of OxfordDPhil

From the biography and OpenAlex affiliations (the years of papers that list the organization), so some years are approximate.

Most-cited works

  1. WhisperX: Time-Accurate Speech Transcription of Long-Form Audio2023 · Interspeech (USB) · 331 citationsWith Andrew Zisserman
  2. Automated audiovisual behavior recognition in wild primates2021 · Science Advances · 86 citationsWith Andrew Zisserman
  3. Frozen in Time: A Joint Video and Image Encoder for End-to-End Retrieval2021 · IEEE/CVF International Conference on Computer Vision (ICCV) · 73 citationsWith Andrew Zisserman
  4. AutoAD: Movie Description in Context2023 · IEEE Computer Society Conference on Computer Vision and Pattern Recognition · 43 citationsWith Andrew Zisserman
  5. A Prompt Array Keeps the Bias Away: Debiasing Vision-Language Models with Adversarial Learning2022 · Conference of the Asia-Pacific Chapter of the Association for Computational Linguistics and the International Joint Conference on Natural Language Processing (AACL-IJCNLP) · 40 citations

From OpenAlex, as of 21 October 2026.

  1. The median h-index is 48 at universities and 14 at frontier AI labsIncludes Max Bain. 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. 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.
  3. Google has the highest combined h-index among AI organizationsIncludes Google. We estimate that Google holds the highest combined h-index at 8,016, while universities account for 73% of the deduplicated total across AI organizations.
  4. 20% of people at companies who shape AI also hold a university role, led by Google and Stanford UniversityIncludes Google. We estimate that 539 of the 6,766 people with a known current organization (8%) hold both a university role and a company role, including 20% of people at companies and 13% of people at universities.

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 Max Bain yet. Send us a free-licensed photo.

  1. Count, Crop and Recognise: Fine-Grained Recognition in the Wild robots.ox.ac.uk. Checked 21 October 2026.
  2. [2005.04208] Condensed Movies: Story Based Retrieval with Contextual Embeddings arxiv.org. Checked 21 October 2026.
  3. [2104.00650] Frozen in Time: A Joint Video and Image Encoder for End-to-End Retrieval arxiv.org. Checked 21 October 2026.
  4. Artificial intelligence used to recognise primate behaviours in the wild eng.ox.ac.uk. Checked 21 October 2026.
  5. [2303.00747] WhisperX: Time-Accurate Speech Transcription of Long-Form Audio arxiv.org. Checked 21 October 2026.
  6. [2307.09006] OxfordVGG Submission to the EGO4D AV Transcription Challenge arxiv.org. Checked 21 October 2026.
  7. [2303.16899] AutoAD: Movie Description in Context arxiv.org. Checked 21 October 2026.
  8. Understanding video through the lens of language - ORA - Oxford University Research Archive ora.ox.ac.uk. Checked 21 October 2026.
  9. Max Bain maxbain.com. Checked 21 October 2026.
  10. [2404.12387] Reka Core, Flash, and Edge: A Series of Powerful Multimodal Language Models arxiv.org. Checked 21 October 2026.
  11. [2405.02287] Vibe-Eval: A hard evaluation suite for measuring progress of multimodal language models arxiv.org. Checked 21 October 2026.
  12. [2507.06261] Gemini 2.5: Pushing the Frontier with Advanced Reasoning, Multimodality, Long Context, and Next Generation Agentic Capabilities arxiv.org. Checked 21 October 2026.
  13. [2608.31170] Context-Aware Interleaved Batching for WhisperX arxiv.org. Checked 21 October 2026.
  14. [2610.05293] BossouChimpanzee: Long-term Chimpanzee Video Dataset arxiv.org. Checked 21 October 2026.