Zemin Wu
Papers
1
Total Citations
38
H-Index
1
About
Zemin Wu is a leading researcher in computer vision, with a primary focus on person re-identification—a critical task for surveillance and security systems. His work bridges the gap between feature engineering and metric learning, challenging the prevailing notion that metric approaches alone hold the advantage. In his highly cited 2015 paper, "Efficient person re-identification by hybrid spatiogram and covariance descriptor" (38 citations), Wu demonstrates that carefully designed features can still achieve state-of-the-art performance. By introducing a hybrid descriptor that fuses spatiogram and covariance information, he provides a computationally efficient yet powerful solution for matching individuals across non-overlapping camera views. This contribution has influenced subsequent research in feature representation for re-identification, inspiring new hybrid approaches. Wu’s work underscores the enduring value of thoughtful feature design in an era dominated by deep learning, making him a respected voice in the field. His research continues to impact both academic studies and practical applications in intelligent video surveillance.
Research Focus
Key Achievements
Top Papers
- 1