Gongyu Zhang

Papers

Total Citations

H-Index

About

Gongyu Zhang is a rising researcher in computer vision and medical image analysis, with a focused interest in surgical instrument segmentation and motion perception. His work addresses the critical challenge of accurately tracking surgical tools in low-quality video feeds, a key bottleneck for autonomous and robot-assisted surgery. In his most-cited paper, "Motion-Boundary-Driven Unsupervised Surgical Instrument Segmentation in Low-Quality Optical Flow" (2025), Zhang introduces a novel unsupervised method that leverages motion boundaries—rather than relying on high-quality optical flow—to segment surgical instruments without the need for labor-intensive annotations. This approach significantly improves robustness in real-world surgical environments where video quality is often degraded by noise, occlusions, and rapid motion. While his citation count is still growing, Zhang’s contribution is notable for its practical relevance: by enabling reliable segmentation in challenging conditions, his work paves the way for safer, more autonomous surgical systems. His research sits at the intersection of low-level vision, optical flow estimation, and medical robotics, offering a fresh perspective on how to handle imperfect data in high-stakes applications.