AI People Wiki

Builders & open source

Trevor Darrell

Faculty, CS and EE Divisions of EECS; founding co-director of BAIR, UC Berkeley

Trevor Darrell is on the faculty of the CS and EE Divisions of the EECS Department at UC Berkeley, where Darrell founded and co-leads the Berkeley Artificial Intelligence Research (BAIR) lab, the Berkeley DeepDrive industrial consortia, and the BAIR Commons program. Darrell's group develops algorithms for large-scale perceptual learning, including object and activity recognition and detection. Darrell was previously on the MIT EECS faculty (1999–2008), directing the Vision Interface Group, and co-founded and serves as President of Prompt AI. Darrell is a co-author of the 2014 Caffe paper — Caffe is a BSD-licensed framework developed by BAIR and the Berkeley Vision and Learning Center with community contributors — and of the R-CNN and ConvNeXt papers.12345

At a glance

Current roles
Faculty, CS and EE Divisions of EECS; founding co-director of BAIR +1 more
Sector
University
Education
MIT, UC Berkeley, Penn
Work listed
3
Sources cited
6
Profile reviewed

Career

Trevor Darrell was a member of the research staff at Interval Research Corporation from 1996 to 1999, then joined the MIT EECS faculty, where Darrell directed the Vision Interface Group until 2008. From 2008 to 2014 Darrell led the Vision group at the International Computer Science Institute, which is affiliated with the University of California.1

At UC Berkeley, Darrell is a founding co-director of Berkeley Artificial Intelligence Research (BAIR), Berkeley DeepDrive and the BAIR Commons program, and was Faculty Director of the PATH research center from 2015 to 2021. Darrell teaches graduate courses on large-scale vision and language models and on moving from research to startup, and co-founded and serves as President of Prompt AI.1

Research

Darrell's group develops algorithms for large-scale perceptual learning, including object and activity recognition and detection, with applications in autonomous vehicles, media search and multimodal interaction with robots and mobile devices. Darrell's interests span computer vision, machine learning, natural language processing and perception-based human-computer interfaces.1

Darrell is a co-author of R-CNN (2013), which combined region proposals with convolutional network features for object detection, and of the 2014 paper on fully convolutional networks for semantic segmentation. Later co-authored work includes "A ConvNet for the 2020s" (2022), which introduced the ConvNeXt family of convolutional models.465

Open-source work

Darrell co-authored the 2014 paper describing Caffe, a deep learning framework. The Caffe repository describes the project as developed by Berkeley AI Research and the Berkeley Vision and Learning Center with community contributors, and releases it under the BSD 2-Clause license.23

Work and contributions

  • Caffe Open-source code

    Co-author of "Caffe: Convolutional Architecture for Fast Feature Embedding" (2014); the BSD-licensed Caffe repository says the framework is developed by BAIR/BVLC and community contributors.23

  • R-CNN Published research

    Co-author, with Ross Girshick, Jeff Donahue, and Jitendra Malik, of "Rich feature hierarchies for accurate object detection and semantic segmentation" (2013), the R-CNN paper.4

  • Fully Convolutional Networks for Semantic Segmentation Published research

    Co-author, with Jonathan Long and Evan Shelhamer, of the 2014 paper on fully convolutional networks for semantic segmentation.6

Timeline

  1. Joined the research staff of Interval Research Corporation, staying until 1999.1
  2. Joined the MIT EECS faculty and directed the Vision Interface Group until 2008.1
  3. Began leading the Vision group at the International Computer Science Institute, through 2014.1
  4. Co-authored the R-CNN paper on rich feature hierarchies for object detection.4
  5. Co-authored the paper introducing the Caffe deep learning framework.2
  6. Co-authored "Fully Convolutional Networks for Semantic Segmentation."6
  7. Became Faculty Director of the PATH research center at UC Berkeley, serving until 2021.1
  8. Co-authored "A ConvNet for the 2020s," which introduced ConvNeXt.5

Other documented roles

Roles the sources document that are not confirmed as current.

  • Faculty, MIT EECS (1999–2008), directing the Vision Interface Group, Massachusetts Institute of Technology1
  • Faculty Director (2015–2021), PATH research center, UC Berkeley1
  • Led the Vision group (2008–2014), International Computer Science Institute1
  • Member of the research staff (1996–1999), Interval Research Corporation1

Where they work

Suggest a correction or removal

Profiles cover public professional work only, and each statement cites a source. If something is out of date, missing or wrong, tell us and link a source. People can also ask for their profile to be corrected or removed.

References

6 sources · profile reviewed Oct 21, 2026 · data updated Oct 21, 2026

  1. Prof. Trevor Darrell — homepage (UC Berkeley EECS)Trevor Darrell · Official page · accessed Oct 21, 2026
  2. Caffe: Convolutional Architecture for Fast Feature Embedding (arXiv 1408.5093)arXiv · Paper · published Jun 20, 2014 · accessed Oct 21, 2026
  3. BVLC/caffe README (developers and BSD license)BVLC (GitHub repository) · Repository · accessed Oct 21, 2026
  4. Rich feature hierarchies for accurate object detection and semantic segmentation (arXiv 1311.2524)arXiv · Paper · published Nov 11, 2013 · accessed Oct 21, 2026
  5. A ConvNet for the 2020s (arXiv 2201.03545)arXiv · Paper · published Jan 10, 2022 · accessed Oct 21, 2026
  6. Fully Convolutional Networks for Semantic Segmentation (arXiv 1411.4038)arXiv · Paper · published Nov 14, 2014 · accessed Oct 21, 2026