Shizhen Zheng
Papers
3
Total Citations
57
H-Index
3
About
Shizhen Zheng is a leading researcher in computer vision and robotics, specializing in stereo vision for indoor scene understanding. Zheng’s major contribution is the creation of the IRS (Indoor Robotics Stereo) dataset, a groundbreaking resource that provides large-scale, naturalistic, and synthetic stereo imagery specifically designed to train deep learning models for disparity and surface normal estimation. This work directly addresses a critical bottleneck in indoor robotics—the need for accurate 3D geometric information to enable reliable localization, navigation, and interaction. The flagship paper on the naturalistic IRS dataset has garnered 31 citations, while the synthetic version has accumulated 16, underscoring the community’s recognition of its value. By bridging the gap between synthetic training data and real-world performance, Zheng’s datasets have become essential benchmarks for advancing stereo-based depth perception. This research is pivotal for developing robust robotic systems that can operate in complex, cluttered indoor environments, moving beyond the limitations of monocular vision. Zheng’s work stands as a cornerstone for anyone building intelligent, spatially-aware robots.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3