Ramakant Nevatia
Ramakant Nevatia, also named Ram Nevatia in his papers, works on computer vision at USC. He earned an electrical engineering doctorate at Stanford in 1975. His laboratory studies shape analysis, object recognition, tracking people and event recognition. With Pramod Sharma, he developed a 2013 method for adapting object detectors to a new video. With Chen Sun, he developed ACTIVE, which uses changes in activity concepts to classify video events. He is Fletcher Jones Professor in Computer Science at USC.
- Works
- 188
- Citations
- 12,065
- h-index
- 35
Publication numbers from OpenAlex, as of 14 October 2026.
Timeline
-
1975
Computer vision training
Nevatia earned an electrical engineering doctorate at Stanford in 1975. He specialized in computer vision for his doctorate. His later work includes shape description, object recognition, stereo analysis and aerial imagery.
-
June 2013
Adapting a detector to a video
Nevatia and Pramod Sharma developed a method for adapting an existing object detector to a new test video. It collects samples without manual labels and trains a random fern classifier to check the baseline detector’s responses. The CVPR paper tests the method for detecting people with two baseline classifiers.
-
December 2013
Recognizing video events
With Chen Sun, Nevatia introduced ACTIVE for classifying high-level events in videos. The method treats short clips as observations in a hidden Markov model and encodes changes in activity concepts with Fisher kernel techniques. Their ICCV paper reports tests on the TRECVID Multimedia Event Detection dataset.
-
As of October 2026
A continuing vision research program
Nevatia is Fletcher Jones Professor in Computer Science, with appointments in computer science and electrical and computer engineering at USC. His laboratory researches tracking humans and recognizing events alongside other vision tasks.
Sources and credits
Photo: viterbi.usc.edu; creator not stated. License: Not stated by source. Source: viterbi.usc.edu. Cropped to a square and resized. Ask us to remove or replace this photo.