Amir Zamir
Amir Zamir studies computer vision and multimodal learning. His group develops perception models that can support agents acting in the real world. Zamir worked at Berkeley, Stanford and UCF before EPFL. His collaborators include Silvio Savarese, Jitendra Malik, Mubarak Shah, Rahul Sukthankar and Leonidas Guibas. His research focuses on general multimodal and multitask perception. His joint projects include Taskonomy, Gibson, 4M and MultiMAE as examples of this program. He coauthored VideoFlexTok at ICML 2026. The project concerns flexible-length, coarse-to-fine video tokenization.
Timeline
-
2026
Flexible video representations
Zamir coauthored VideoFlexTok at ICML 2026. The project concerns flexible-length, coarse-to-fine video tokenization.
-
As of October 2026
Research path
Zamir worked at Berkeley, Stanford and UCF before EPFL. His collaborators include Silvio Savarese, Jitendra Malik, Mubarak Shah, Rahul Sukthankar and Leonidas Guibas.
-
As of October 2026
Learning across tasks and modalities
His research focuses on general multimodal and multitask perception. His joint projects include Taskonomy, Gibson, 4M and MultiMAE as examples of this program.
Sources and credits
Photo: ellis.eu; Amir Zamir profile photograph; creator not stated. License: Not stated by source. Source: ellis.eu. Cropped to a square and resized. Ask us to remove or replace this photo.