Huizong Feng

Chongqing University of Posts and Telecommunications

Papers

Total Citations

H-Index

About

Dr. Huizong Feng is a rising researcher at the intersection of computer vision and autonomous systems, with a primary focus on visual odometry (VO) and semantic scene understanding. His most cited work, "Visual Odometry with Deep Learning for Joint Semantic Segmentation" (2024), tackles a critical challenge in motion estimation for autonomous vehicles, robots, and augmented reality: the scale ambiguity and large relative pose errors that plague existing learning-based VO methods. By integrating deep learning with semantic segmentation, Feng proposes a framework that jointly estimates camera motion while understanding scene context, improving both accuracy and robustness in dynamic environments. Though early in his career—with his top paper currently accumulating 1 citation—his work addresses a foundational problem in embodied AI, where precise ego-motion estimation is essential for safe navigation. Feng’s research sits at the nexus of deep learning, 3D geometry, and semantic reasoning, promising to advance the reliability of autonomous systems. As the field increasingly demands models that perceive both motion and meaning, Feng’s contributions position him as a promising voice in next-generation visual perception.