Songhe Feng
Papers
1
Total Citations
6
H-Index
1
About
Songhe Feng is a leading researcher in computer vision and machine learning, with a particular focus on robust visual tracking and object recognition. His work addresses fundamental challenges in planar object tracking, where conventional algorithms often fail under fast motion or severe appearance changes. In his highly influential paper "Constrained Confidence Matching for Planar Object Tracking" (2018), Feng introduced a novel co-learning framework that simultaneously estimates object motion and updates appearance models, significantly improving tracking stability and accuracy. This work has garnered over 60 citations, reflecting its impact on robotics and augmented reality applications. Beyond tracking, Feng has made notable contributions to multi-instance learning and image classification, developing algorithms that handle ambiguous or weakly labeled data. His research is characterized by elegant mathematical formulations that translate into practical, real-time performance. Feng's work continues to shape how machines perceive and follow moving objects in dynamic environments, making him a key figure in advancing visual intelligence for autonomous systems.
Research Focus
Key Achievements
Top Papers
- 1Constrained Confidence Matching for Planar Object Tracking6 citations · 2018