Papers
4
Total Citations
82
H-Index
3
About
Kaite Xiang is a computer vision researcher specializing in semantic segmentation, autonomous systems, and assistive navigation technologies. Her work sits at the intersection of deep learning and real-world perception challenges, with a particular focus on extending the capabilities of scene understanding beyond conventional camera constraints. Xiang's most significant contribution is the development of DS-PASS (Detail-Sensitive Panoramic Annular Semantic Segmentation), introduced through her SwaftNet architecture, which addresses a critical gap in autonomous driving and robotics research: the dominance of narrow field-of-view pipelines designed for standard pinhole cameras. By enabling high-quality semantic segmentation across panoramic annular images, her work substantially broadens the perceptual horizon available to autonomous transportation systems, garnering over 60 citations across its publications. Complementing this, her Importance-Aware Semantic Segmentation framework introduces an efficient pyramidal context network that prioritizes semantically critical objects — a meaningful advance for navigational assistant systems serving vulnerable road users. With a research portfolio accumulating over 80 citations, Xiang has demonstrated consistent influence in making autonomous and assistive systems more robust, context-aware, and practically deployable. Her contributions are particularly valuable for researchers working on wide-angle perception, real-time scene understanding, and intelligent mobility solutions.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3
- 4