Yazhou Liu
Papers
1
Total Citations
3
H-Index
1
About
Yazhou Liu is a computer vision researcher whose work focuses on advancing human detection and image analysis through innovative feature descriptors. His key contributions lie in developing granularity-tunable gradient-based representations, most notably the Isotropic Granularity-tunable Gradients Partition (IGGP) descriptor introduced in his 2010 paper. This work extended existing GGP descriptors by achieving isotropic feature alignment, enabling more robust human detection in still images. While his most-cited paper has garnered 3 citations, Liu’s research addresses fundamental challenges in visual recognition—specifically, how to create feature representations that are both granularity-adaptive and orientation-invariant. His approach to partitioning gradient information at multiple scales has implications for pedestrian detection, surveillance, and autonomous systems. Liu’s work contributes to the broader effort in computer vision to balance descriptive power with computational efficiency, making his descriptors suitable for real-world applications where viewpoint and scale variations are common. His research continues to influence the development of feature engineering techniques for object detection and image understanding.
Research Focus
Key Achievements
Top Papers
- 1