Benjin Zhu
Papers
3
Total Citations
94
H-Index
3
About
Benjin Zhu is a rising star in 3D computer vision, with a focused research program advancing perception for autonomous driving and robotics. His core contributions lie in temporal modeling for point cloud sequences, addressing the critical challenge of detecting and tracking objects reliably over time. Zhu is best known for introducing **MPPNet**, a flexible and high-performance framework for 3D temporal object detection that uses novel "proxy points" to intertwine multi-frame features, achieving state-of-the-art results (garnering 79 citations for its primary publication). Building on this, he developed **TrajectoryFormer**, a transformer-based approach for 3D multi-object tracking that predicts trajectory hypotheses to overcome the limitations of standard tracking-by-detection paradigms. His work directly tackles the need for robust, real-time perception in dynamic environments. With over 90 total citations, Zhu's research is already influencing the design of safer autonomous systems, and his innovative use of temporal context marks him as a key contributor to the next generation of 3D scene understanding.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3