Fan Xuanzhe
Papers
1
Total Citations
8
H-Index
1
About
Fan Xuanzhe is a rising researcher in computer vision and human-robot interaction, with a focused expertise in 3D human motion prediction. His most notable contribution is the development of a Deformable Transformer-based Adversarial Network, a novel architecture that addresses two critical challenges in the field: the tendency of transformer models to collapse into non-plausible poses and their prohibitive quadratic computational complexity. By integrating deformable attention mechanisms with adversarial training, Xuanzhe's work achieves both high-fidelity motion generation and significantly improved efficiency, making real-time human-robot interaction more feasible. His 2022 paper on this method has already garnered 8 citations, signaling its growing impact among peers. This work represents a meaningful step toward more natural and responsive robotic systems, positioning Xuanzhe as a promising voice in the intersection of deep learning and embodied AI. His research continues to push the boundaries of how machines understand and anticipate human movement.
Research Focus
Key Achievements
Top Papers
- 1