Bian Xihan
Papers
2
Total Citations
9
H-Index
2
About
Bian Xihan is a rising researcher in robot learning, whose work sits at the intersection of imitation learning, reinforcement learning, and multi-task transfer. Her research addresses a fundamental challenge in robotics: enabling machines to flexibly adapt learned behaviours across different tasks and environments, much like humans do. In her highly cited work *SKILL-IL: Disentangling Skill and Knowledge in Multitask Imitation Learning* (5 citations), she introduced a novel framework that separates transferable skills from task-specific knowledge, allowing a robot to repurpose its abilities—for example, applying a cycling skill to a driving context. Her earlier paper, *Robot in a China Shop: Using Reinforcement Learning for Location-Specific Navigation Behaviour* (4 citations), reframes navigation as a multi-task learning problem, enabling robots to deploy distinct, context-appropriate behaviours in different settings, such as a cluttered shop versus an open warehouse. Though early in her career, Bian’s work is already shaping how researchers think about scalable, generalist robot intelligence. Her contributions are particularly notable for their elegant blend of cognitive inspiration and algorithmic rigour, laying groundwork for robots that can truly learn and adapt across the messy, varied environments of the real world.
Research Focus
Key Achievements
Top Papers
- 1
- 2