Alexander Berg
Also known as Alex Berg, Alexander Christiansen Berg, Alexander C. Berg
Alexander Berg is a computer-vision researcher and associate professor at UC Irvine. His work spans image recognition, image description, and promptable segmentation. He studied mathematics at Johns Hopkins, completed a computer-science doctorate at Berkeley, and held research and faculty roles at Yahoo, Columbia, Stony Brook, and UNC Chapel Hill. His coauthored work on image descriptions and entry-level categories received the Longuet-Higgins and Marr prizes; his computer-vision work also received the Everingham Prize. At UCI since 2022, he has coauthored Segment Anything and research on text-to-image camera control; in 2026, he and the SSD team received the Koenderink Prize.
- Works
- 114
- Citations
- 69,388
- h-index
- 49
Publication numbers from OpenAlex, as of 14 October 2026.
Timeline
-
1994–2005
From mathematics to AI
He earned bachelor's and master's degrees in mathematics at Johns Hopkins in 1994, then completed a computer-science PhD at UC Berkeley in 2005.
-
2011
Teaching images to speak
In Baby Talk, he and his coauthors used computer-vision recognition and text statistics to generate simple natural-language image descriptions.
-
2013
Giving objects everyday names
By August 2013, he was an assistant professor at UNC Chapel Hill after an earlier post at Stony Brook. His ICCV paper modeled entry-level categories, the everyday labels people use for objects in images, and won the Marr Prize.
-
2016
Scaling image recognition
He and colleagues developed the ImageNet dataset and competition. His computer-vision work received the Everingham Prize for community contributions in 2016.
-
2021
A paper earns its prize
CVPR's 2021 Longuet-Higgins Prize recognized Baby Talk, his 2011 paper on simple image descriptions.
-
By 2022
A career across vision labs
His career includes research-scientist roles at Columbia University and Yahoo! Research, faculty posts at Stony Brook and UNC Chapel Hill, work in FAIR at Meta, and cofounding Shopagon. He became an associate professor at UC Irvine in March 2022.
-
Since 2022
Prompting computer vision
He coauthored Segment Anything, a promptable image-segmentation model, and a CVPR paper on camera control for text-to-image generation.
-
2026
SSD passes a time test
He and the SSD team received the 2026 Koenderink Prize at ECCV for SSD: Single Shot Multibox Detector.
Sources and credits
Photo: vision.ics.uci.edu; Alexander Berg | Computational Vision | UC Irvine; creator not stated. License: Not stated by source. Source: vision.ics.uci.edu. Cropped to a square and resized. Ask us to remove or replace this photo.
- Alex Berg - UC Irvine Donald Bren School of Information & Computer Sciences
- Baby Talk: Understanding and Generating Simple Image Descriptions | ML Anthology
- Alex Berg Computer Vision UNC Chapel Hill
- From large scale image categorization to entry-level categories - Princeton University
- Awards - iccv2013
- AI at ICS: Six Decades of AI (History / Overview) - UC Irvine Donald Bren School of Information & Computer Sciences
- CS 201: Computation and Vision, ALEX BERG, University of North Carolina at Chapel Hill | CS
- Yejin Choi
- Alexander Christiansen Berg aka Alex Berg
- Alex Berg Computer Vision UC Irvine
- Segment Anything
- Alexander Berg | Computational Vision | UC Irvine
- News & Awards · CML at UC Irvine