Komeili
Papers
1
Total Citations
3
H-Index
1
About
Mojtaba Komeili is a leading researcher in self-supervised learning, video understanding, and robotics, whose work bridges the gap between passive observation and active interaction in artificial intelligence. His most notable contribution is the development of V-JEPA 2, a self-supervised video model that learns to understand, predict, and plan by leveraging internet-scale video data alongside minimal interaction data from robot trajectories. This approach addresses a fundamental challenge in AI: enabling models to comprehend the world largely through observation, with only a small amount of hands-on experience. While still early in its impact, with 3 citations since 2025, V-JEPA 2 represents a paradigm shift toward more efficient and generalizable AI systems that can learn from abundant, unlabeled video rather than requiring extensive manual annotation or interaction. Komeili’s work sits at the intersection of computer vision, representation learning, and embodied AI, offering a path toward machines that can perceive and act in the world with minimal supervision—a key step toward more autonomous and adaptable intelligent systems.
Research Focus
Key Achievements
Top Papers
- 1