Nikita Dhawan
Papers
1
Total Citations
31
H-Index
1
About
Nikita Dhawan is a researcher at the forefront of robot learning from human demonstrations, with a focus on enabling robots to perform complex, multi-stage tasks. Her most-cited work, "AVID: Learning Multi-Stage Tasks via Pixel-Level Translation of Human Videos" (2020, 31 citations), introduces a novel framework that translates human instructional videos into robot-executable actions using pixel-level translation with CycleGAN. This approach allows robots to learn task stages directly from human examples, automatically resetting and retrying until each stage is successfully completed—a significant leap toward more intuitive and efficient robot training. By bridging the gap between human and robot action spaces at the visual level, Dhawan’s work reduces the need for costly manual programming or extensive robot-specific data collection. Her research has important implications for robotics, human-robot interaction, and imitation learning, offering a scalable pathway for teaching robots everyday manipulation tasks. As a rising scholar, Dhawan’s contributions are shaping how robots can learn from and adapt to human guidance in real-world settings.
Research Focus
Key Achievements
Top Papers
- 1AVID: Learning Multi-Stage Tasks via Pixel-Level Translation of Human Videos31 citations · 2020