PeopleofSuperintelligence

Researcher

Matthijs Douze

Matthijs Douze works on computer vision and similarity search, the methods that let software find items that resemble a query. After a PhD in computer vision at the University of Toulouse, he joined INRIA's LEAR team in Grenoble. Faiss, a library for similarity search, was introduced in a 2017 Meta engineering post that he co-wrote. The Hamming Embedding paper he co-wrote received the Koenderink Prize at ECCV 2018. As of October 2026, he is listed under Meta FAIR on a May 2026 paper.

Works
223
Citations
24,889
h-index
44

Publication numbers from OpenAlex, as of 15 October 2026.

Timeline

  1. By 2010

    Research engineer at INRIA

    He earns a PhD in computer vision at the University of Toulouse, joins INRIA's LEAR team in Grenoble as a research engineer, and in 2010 becomes a permanent member of the SED service.

  2. November 2015

    On leave to Facebook AI Research

    He goes on leave to Facebook AI Research in mid-November 2015.

  3. March 2017

    Releasing the Faiss library

    Hervé Jegou, Matthijs Douze and Jeff Johnson publish a Meta engineering post in March 2017, introducing Faiss, a library for quickly searching multimedia documents that are similar to each other.

  4. 2018

    A Koenderink Prize

    At ECCV 2018, the Koenderink Prize goes to the Hamming Embedding paper by Hervé Jegou, Matthijs Douze and Cordelia Schmid.

  5. 2024

    The Faiss library paper

    Among the authors of the 2024 arXiv paper The Faiss library, he is listed first, followed by Alexandr Guzhva, Chengqi Deng and Jeff Johnson.

  6. As of October 2026

    Co-authoring at Meta FAIR

    He is listed under Meta FAIR on the May 2026 paper Self-Pruned Key-Value Attention, and he co-authors Sparse Delta Memory in July 2026.

Sources and credits

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

  1. Matthijs Douze thoth.inrialpes.fr. Checked 15 October 2026.
  2. Faiss: A library for efficient similarity search - Engineering at Meta engineering.fb.com. Checked 15 October 2026.
  3. ECVA | European Computer Vision Association ecva.net. Checked 15 October 2026.
  4. [2401.08281] The Faiss library arxiv.org. Checked 15 October 2026.
  5. [2605.14037] Self-Pruned Key-Value Attention: Learning When to Write by Predicting Future Utility arxiv.org. Checked 15 October 2026.
  6. [2607.07386] Sparse Delta Memory: Scaling the State of Linear RNNs through Sparsity arxiv.org. Checked 15 October 2026.