Abha Gejji
Papers
1
Total Citations
3
H-Index
1
About
Abha Gejji is a leading researcher at the forefront of self-supervised learning and video understanding, whose work is redefining how AI systems perceive and interact with the world. Her primary research areas span video representation learning, world modeling, and the intersection of computer vision with robotics. Gejji’s most notable contribution is her work on the V-JEPA 2 framework, a self-supervised video model that learns to understand, predict, and plan by observing internet-scale video data, requiring only minimal interaction data from robot trajectories. This groundbreaking approach addresses a fundamental challenge in modern AI: enabling machines to grasp the dynamics of the physical world largely through passive observation. With 3 citations already in its first year, V-JEPA 2 is quickly gaining recognition for its potential to bridge the gap between perception and action. Gejji’s research is particularly impactful for students and researchers interested in building more efficient, generalizable AI systems that can learn from abundant, unlabeled video rather than relying on costly labeled datasets or extensive real-world interaction. Her work is paving the way for more autonomous robots and intelligent systems that can plan and act in complex environments.
Research Focus
Key Achievements
Top Papers
- 1