Alexei A. Efros
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Alexei A. Efros has a total of 27 co-authors. Here are the top co-authors by number of shared publications:
- Alexander C. Li
Carnegie Mellon University
2 shared publications - Deepak Pathak
Carnegie Mellon University
2 shared publications - Martial Hebert
Carnegie Mellon University
2 shared publications - Mathieu Aubry
LIGM, Ecole des Ponts, Univ Gustave Eiffel, CNRS
2 shared publications - Shiry Ginosar
TTIC, Chicago, USA
2 shared publications - Aleksander Holynski
University of California, Berkeley, San Francisco, United States of America and Google DeepMind, San Francisco, United States of America
1 shared publication
Latest Publications
- 11 September 2026
- Published by Springer Nature
- p. 332-350
Bottom-Up Modeling of Repeated Elements via Single Image Analysis-by-Synthesis
- 11 September 2026
- Published by Springer Nature
- p. 332-350
Bottom-Up Modeling of Repeated Elements via Single Image Analysis-by-Synthesis
- Open Access
- 19 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-11
Autoregressive Modeling of Film with Applications in Video Montage
- 19 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-11
Open AccessAutoregressive Modeling of Film with Applications in Video Montage
- 14 February 2023
- Published by Springer Nature
- p. 100-104
Internet Curiosity: Directed Unsupervised Learning on Uncurated Internet Data
- 14 February 2023
- Published by Springer Nature
- p. 100-104
Internet Curiosity: Directed Unsupervised Learning on Uncurated Internet Data
- 23 October 2022
- Published by Springer Nature
- p. 490-505
Understanding Collapse in Non-contrastive Siamese Representation Learning
- 23 October 2022
- Published by Springer Nature
- p. 490-505
Understanding Collapse in Non-contrastive Siamese Representation Learning
- Open Access
Learning Pixel Trajectories with Multiscale Contrastive Random Walks
A range of video modeling tasks, from optical flow to multiple object tracking, share the same fundamental challenge: establishing space-time correspondence. Yet, approaches that dominate each space differ. We take a step towards bridging this gap by extending the recent contrastive random walk formulation to much denser, pixel-level space-time graphs. The main contribution is introducing hierarchy into the search problem by computing the transition matrix between two frames in a coarse-to-fine manner, forming a multiscale contrastive random walk when extended in time. This establishes a unified technique for self-supervised learning of optical flow, keypoint tracking, and video object segmentation. Experiments demonstrate that, for each of these tasks, the unified model achieves performance competitive with strong self-supervised approaches specific to that task. Project webpage: https://jasonbian97.github.io/flowwalk
Open AccessLearning Pixel Trajectories with Multiscale Contrastive Random Walks
A range of video modeling tasks, from optical flow to multiple object tracking, share the same fundamental challenge: establishing space-time correspondence. Yet, approaches that dominate each space differ. We take a step towards bridging this gap by extending the recent contrastive random walk formulation to much denser, pixel-level space-time graphs. The main contribution is introducing hierarchy into the search problem by computing the transition matrix between two frames in a coarse-to-fine manner, forming a multiscale contrastive random walk when extended in time. This establishes a unified technique for self-supervised learning of optical flow, keypoint tracking, and video object segmentation. Experiments demonstrate that, for each of these tasks, the unified model achieves performance competitive with strong self-supervised approaches specific to that task. Project webpage: https://jasonbian97.github.io/flowwalk
- Open Access
Self-Supervised Policy Adaptation during Deployment
In most real world scenarios, a policy trained by reinforcement learning in one environment needs to be deployed in another, potentially quite different environment. However, generalization across different environments is known to be hard. A natural solution would be to keep training after deployment in the new environment, but this cannot be done if the new environment offers no reward signal. Our work explores the use of self-supervision to allow the policy to continue training after deployment without using any rewards. While previous methods explicitly anticipate changes in the new environment, we assume no prior knowledge of those changes yet still obtain significant improvements. Empirical evaluations are performed on diverse simulation environments from DeepMind Control suite and ViZDoom, as well as real robotic manipulation tasks in continuously changing environments, taking observations from an uncalibrated camera. Our method improves generalization in 31 out of 36 environments across various tasks and outperforms domain randomization on a majority of environments.
Open AccessSelf-Supervised Policy Adaptation during Deployment
In most real world scenarios, a policy trained by reinforcement learning in one environment needs to be deployed in another, potentially quite different environment. However, generalization across different environments is known to be hard. A natural solution would be to keep training after deployment in the new environment, but this cannot be done if the new environment offers no reward signal. Our work explores the use of self-supervision to allow the policy to continue training after deployment without using any rewards. While previous methods explicitly anticipate changes in the new environment, we assume no prior knowledge of those changes yet still obtain significant improvements. Empirical evaluations are performed on diverse simulation environments from DeepMind Control suite and ViZDoom, as well as real robotic manipulation tasks in continuously changing environments, taking observations from an uncalibrated camera. Our method improves generalization in 31 out of 36 environments across various tasks and outperforms domain randomization on a majority of environments.
- 6 July 2016
- Published by Springer Nature
- p. 21-40
Discovering Mid-level Visual Connections in Space and Time
- 6 July 2016
- Published by Springer Nature
- p. 21-40
Discovering Mid-level Visual Connections in Space and Time
- Open Access
- 1 September 2009
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- Vol. 23, 681-688
Image composition for object pop-out
We propose a new data-driven framework for novel object detection and segmentation, or ¿object pop-out¿. Traditionally, this task is approached via background subtraction, which requires continuous observation from a stationary camera. Instead, we consider this an image matching problem. We detect novel objects in the scene using an unordered, sparse database of previously captured images of the same general environment. The problem is formulated in a new image composition framework: 1) given an input image, we find a small set of similar matching images; 2) each of the matches is aligned with the input by proposing a set of homography transformations; 3) regions from different transformed matches are stitched together into a single composite image that best matches the input; 4) the difference between the input and the composite is used to ¿pop-out¿ new or changed objects.
- 1 September 2009
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- Vol. 23, 681-688
Open AccessImage composition for object pop-out
We propose a new data-driven framework for novel object detection and segmentation, or ¿object pop-out¿. Traditionally, this task is approached via background subtraction, which requires continuous observation from a stationary camera. Instead, we consider this an image matching problem. We detect novel objects in the scene using an unordered, sparse database of previously captured images of the same general environment. The problem is formulated in a new image composition framework: 1) given an input image, we find a small set of similar matching images; 2) each of the matches is aligned with the input by proposing a set of homography transformations; 3) regions from different transformed matches are stitched together into a single composite image that best matches the input; 4) the difference between the input and the composite is used to ¿pop-out¿ new or changed objects.