PeopleofSuperintelligence

Researcher

Matthias Minderer

Matthias Minderer is a computer-vision researcher whose work spans visual neuroscience, video modeling, image recognition, and open-vocabulary detection. He studied neuroscience at ETH Zürich and biochemistry at Cambridge before completing a Harvard PhD with Christopher Harvey. He coauthored papers on object dynamics and Vision Transformers, and contributed to PaliGemma's spatial understanding. Since 2025, he works at Microsoft AI in Zürich on multimodal large language models focused on spatial understanding and localization.

Now at
Microsoft
Badges

Timeline

  1. 2019

    Harvard PhD and video dynamics

    At Harvard, he earns a PhD while working with Christopher Harvey; his thesis focuses on visual and action-related information in the mammalian cerebral cortex. Earlier, he studies neuroscience at ETH Zürich and biochemistry at the University of Cambridge. In 2019, he coauthors a NeurIPS paper on object structure and dynamics from videos.

  2. 2020

    Removing shortcuts in self-supervised learning

    In 2020, he coauthors a paper on removing shortcuts from self-supervised representation learning. It trains a lens network to make small image changes that reduce performance on a pretext task.

  3. 2021

    Vision Transformers for image recognition

    In 2021, he coauthors a Vision Transformer paper for image recognition. It applies a pure transformer directly to image-patch sequences for image classification.

  4. 2022–2023

    OWL-ViT and OWLv2

    In 2022, he coauthors OWL-ViT, which uses a Vision Transformer with contrastive image-text pretraining for open-vocabulary object detection. In 2023, he coauthors OWLv2, which scales detection data with self-training and uses an existing detector to generate pseudo-box annotations on image-text pairs.

  5. 2024

    PaliGemma spatial understanding

    In 2024, he coauthors PaliGemma, an open 3B vision-language model based on the SigLIP-So400m vision encoder and Gemma-2B language model. He contributes to its spatial-understanding capabilities.

  6. Since 2025

    Microsoft AI in Zürich

    In February 2025, he joins Microsoft AI as a founding member of its new Zürich office. Before the move, he works as a research scientist at Google Brain and later Google DeepMind. He is a Member of Technical Staff focused on multimodal large language models, spatial understanding, and localization.

Career

  1. Since 2025MicrosoftMember of Technical Staff
  2. Google BrainResearch Scientist
  3. Google DeepMindResearch Scientist

Education

  1. 2019Harvard UniversityPhD
  2. ETH Zurich
  3. University of Cambridge

From the biography.

  1. Of 943 people shaping AI who worked at Google, 446 work there nowIncludes Matthias Minderer. 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. 16% of AI career moves in the 2020s went from a university to a companyIncludes Matthias Minderer. 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. Among employers with 30+ full-time cases, OpenAI leads primary technical role H-1B pay at $310,000, and Amazon, Google and Microsoft file 57% of all casesIncludes Microsoft. Across 440,669 certified US H-1B filings from 44 employers (FY2020–FY2026), OpenAI leads primary technical role base pay at $310,000 and xAI leads overall base pay at $290,000.
  4. Google had the most people shaping AI every year from 2014 to 2025Includes Microsoft. 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 with dated positions at the same organization in the same years.

Sources and credits

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

  1. Matthias Minderer matthias.minderer.net. Checked 21 October 2026.
  2. Unsupervised learning of object structure and dynamics from videos proceedings.neurips.cc. Checked 21 October 2026.
  3. Automatic Shortcut Removal for Self-Supervised Representation Learning proceedings.mlr.press. Checked 21 October 2026.
  4. An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale research.google. Checked 21 October 2026.
  5. Simple Open-Vocabulary Object Detection with Vision Transformers research.google. Checked 21 October 2026.
  6. Scaling Open-Vocabulary Object Detection arxiv.org. Checked 21 October 2026.
  7. PaliGemma: A versatile 3B VLM for transfer arxiv.org. Checked 21 October 2026.
  8. Microsoft eröffnet KI-Labor in Zürich netzwoche.ch. Checked 21 October 2026.