Dongxu Zhao

University of North Carolina at Chapel Hill

Papers

Total Citations

H-Index

About

Dongxu Zhao is a rising researcher in computer vision and robotics, whose work focuses on bridging the gap between 2D perception and 3D scene understanding. His key research areas include multi-view geometry, depth estimation, and semantic segmentation—particularly through the lens of foundation models like the Segment Anything Model (SAM). In his highly cited 2024 paper, "Joint Depth Prediction and Semantic Segmentation with Multi-View SAM," Zhao introduced a novel multi-task framework that leverages multiple camera viewpoints to simultaneously predict depth and segment objects, overcoming the inherent limitations of single-view approaches. This work is especially impactful for real-world robotics applications, where multi-view data is readily available but underutilized. By avoiding the computational overhead of full 3D methods while achieving richer scene understanding, Zhao’s contribution has already garnered 10 citations in its first year, signaling strong influence in the community. His research offers a practical, scalable path for autonomous systems to perceive and navigate complex environments, making him a promising voice in the next generation of vision-based robotics.