Guangkun Feng
Papers
1
Total Citations
19
H-Index
1
About
Guangkun Feng is a rising researcher in computer vision and robotics, whose work focuses on advancing 6D pose estimation—a critical task for enabling machines to perceive and interact with objects in three-dimensional space. His most cited paper, "NVR-Net: Normal Vector Guided Regression Network for Disentangled 6D Pose Estimation" (2023, 19 citations), introduces a novel approach that decouples rotation and translation estimation from monocular images. By leveraging normal vector guidance, Feng’s method overcomes limitations of traditional two-stage pipelines like Perspective-n-Point (PnP), which often suffer from accuracy degradation due to entangled pose parameters. This contribution directly addresses a fundamental challenge in augmented reality, autonomous manipulation, and industrial automation. While still early in his career, Feng’s work demonstrates a knack for rethinking established frameworks—his disentangled regression strategy offers a cleaner, more precise alternative to conventional techniques. As the demand for robust object-level perception grows, his research is poised to influence both academic benchmarks and real-world robotic systems. With a focus on geometric reasoning and deep learning integration, Feng represents a new generation of vision researchers bridging theory and practical deployment.
Research Focus
Key Achievements
Top Papers
- 1