Wenzhe Cai
Papers
9
Total Citations
101
H-Index
4
About
Wenzhe Cai is an emerging researcher at the forefront of robotic navigation, reinforcement learning, and human-robot interaction, with work that bridges cutting-edge AI foundation models and real-world robotic systems. His most impactful contribution, "Bridging Zero-shot Object Navigation and Foundation Models through Pixel-Guided Navigation Skill" (36 citations), addresses a critical challenge in home-assistance robotics by integrating the visual grounding and commonsense reasoning capabilities of foundation models with practical robot locomotion skills. His research in multi-task reinforcement learning, particularly his attention-based mixture-of-experts framework (25 citations), advances how robots can efficiently share and leverage knowledge across diverse tasks simultaneously. Cai has also made notable strides in crowd-aware robot navigation with dynamic human preference modeling, UAV target tracking under complex occlusions, and zero-shot instruction-following navigation in unexplored environments. His recent work on affordance-guided prompting for large language models demonstrates a keen interest in grounding AI reasoning within physical reality. Collectively accumulating nearly 100 citations, Cai's research consistently tackles the sim-to-real gap and multi-modal fusion challenges, making him a compelling voice in the rapidly evolving field of intelligent autonomous robotics.
Research Focus
Key Achievements
Top Papers
- 1
- 2Multi-Task Reinforcement Learning With Attention-Based Mixture of Experts25 citations · 2023
- 3
- 4
- 5
- 6
- 7
- 8
- 9Robust Navigation with Cross-Modal Fusion and Knowledge Transfer2 citations · 2023