Papers
2
Total Citations
11
H-Index
2
About
Sui Wei is a rising researcher in computer vision and robotics, specializing in transparent object perception and geometric scene understanding. Their work addresses fundamental challenges in enabling machines to interact with visually complex environments. In their highly cited 2025 paper, "Monocular Depth Estimation and Segmentation for Transparent Object with Iterative Semantic and Geometric Fusion" (8 citations), Sui Wei tackles the notoriously difficult problem of perceiving transparent objects—which lack distinct visual features—by developing a novel iterative fusion framework that simultaneously estimates depth and performs segmentation without requiring specialized sensors. This contribution is critical for robotic manipulation tasks involving glass, plastic, or liquids. Earlier, in "Sitpose: A Siamese Convolutional Transformer for Relative Camera Pose Estimation" (2023, 3 citations), they introduced an innovative siamese convolutional transformer model that directly regresses relative camera pose from overlapping image pairs, advancing visual odometry and 3D reconstruction. By combining convolutional and transformer architectures, Sui Wei demonstrates a talent for hybrid models that leverage the strengths of both paradigms. Their work has immediate applications in autonomous navigation, augmented reality, and industrial automation, establishing them as a promising voice in perception for robotics.
Research Focus
Key Achievements
Top Papers
- 1
- 2