Papers
8
Total Citations
461
H-Index
6
About
Yuankai Qi is a prominent researcher specializing in Vision-and-Language Navigation (VLN) and embodied AI, with a focus on enabling robots and autonomous agents to understand and act upon natural language instructions within real visual environments. His most influential contribution is the REVERIE benchmark — Remote Embodied Visual Referring Expression in Real Indoor Environments — which has garnered nearly 300 citations since its 2020 publication and has become a foundational challenge in the robotics and AI communities. REVERIE distinguishes itself by presenting agents with high-level, human-like instructions rather than step-by-step guidance, pushing the frontier of autonomous navigation research. Qi further advanced the field through his Object-and-Action Aware Model, which addresses the dual challenge of extracting meaningful object references and action directives from natural language to guide agent behavior. His more recent works, including "March in Chat" and "Mind the Gap," demonstrate a continued commitment to refining navigation success through interactive prompting and bridging critical performance gaps in existing VLN systems. Collectively, Qi's research has accumulated over 450 citations, underscoring his significant and growing impact on the intersection of computer vision, natural language processing, and embodied intelligence.
Research Focus
Key Achievements
Top Papers
- 1REVERIE: Remote Embodied Visual Referring Expression in Real Indoor Environments297 citations · 2020
- 2Object-and-Action Aware Model for Visual Language Navigation98 citations · 2020
- 3
- 4
- 5
- 6Object-and-Action Aware Model for Visual Language Navigation7 citations · 2020
- 7
- 8