Alexei A. Efros
Also known as Alexei (Alyosha) Efros, Alexei Efros, Alyosha Efros
Alexei A. Efros develops data-driven methods for computer vision, graphics and computational photography. Efros earned a Berkeley doctorate in October 2003, held an Oxford research fellowship and worked at CMU from 2004 to 2013 before joining Berkeley. His research uses large amounts of unlabeled visual data to understand and recreate visual scenes. He coauthored pix2pix, which studies image-to-image translation with conditional adversarial networks. He also works on self-supervised learning, world models and continual learning.
- Works
- 257
- Citations
- 116,483
- h-index
- 100
Publication numbers from OpenAlex, as of 14 October 2026.
Timeline
-
1997–2003
Computer-science training
Efros earned a Utah bachelor’s degree in June 1997, a Berkeley master’s degree in December 1999 and a Berkeley doctorate in October 2003. His thesis studied data-driven texture and motion, supervised by Jitendra Malik.
-
2003–2013
Oxford and Carnegie Mellon
Efros held an Oxford visiting research fellowship from 2003 to 2004, then a CMU assistant professorship from 2004 to 2010 and an associate professorship from 2010 to 2013.
-
2013–2017
Berkeley and image translation
Efros became a Berkeley associate professor in 2013 and professor in 2017. In November 2016 he coauthored the pix2pix preprint with Phillip Isola, Jun-Yan Zhu and Tinghui Zhou. The method learns conditional mappings between images.
-
As of October 2026
Learning from visual data
Efros works on data-driven vision, image generation, self-supervised learning, continual learning and world models. He is a Berkeley EECS professor and is affiliated with BAIR.
Sources and credits
Photo: www2.eecs.berkeley.edu; Alexei (Alyosha) Efros | EECS at UC Berkeley; creator not stated. License: Not stated by source. Source: www2.eecs.berkeley.edu. Cropped to a square and resized. Ask us to remove or replace this photo.