Deva Ramanan
Deva Ramanan researches computer vision and machine learning at Carnegie Mellon’s Robotics Institute. His research emphasizes temporal information in perception and learning from visual observations. His work includes multiscale deformable part models and multi-class object layout, as well as the COCO dataset. He received research recognition and served as program chair of CVPR 2018. His work connects recognition algorithms, datasets and the study of how visual systems learn useful regularities.
- Works
- 305
- Citations
- 84,055
- h-index
- 83
Publication numbers from OpenAlex, as of 14 October 2026.
Timeline
-
2009
Models of object layout
Chaitanya Desai, Ramanan and Charless Fowlkes coauthored the 2009 paper on discriminative models for multi-class object layout.
-
2018
Part models and conference service
The multiscale deformable-part-model paper by Pedro Felzenszwalb, David McAllester and Ramanan received CVF’s 2018 test-of-time prize. He served as CVPR program chair that year.
-
2024
COCO test-of-time award
CVF gave a 2024 test-of-time award to Microsoft COCO: Common Objects in Context, with Ramanan among the coauthors.
-
As of October 2026
Temporal and data-driven perception
He is a professor at CMU’s Robotics Institute. He conducts vision research using machine learning and human perception as inspiration, including the value of temporal information and the challenge of learning from few examples.
Sources and credits
Photo: www.cs.cmu.edu; creator not stated. License: Not stated by source. Source: cs.cmu.edu. Cropped to a square and resized. Ask us to remove or replace this photo.