Papers
3
Total Citations
45
H-Index
3
About
Kai Berger is a researcher at the intersection of computer vision and robotics, with a primary focus on 3-D perception in challenging, real-world environments. His key research areas include stereo vision, polarization imaging, and scene understanding for autonomous systems, particularly in urban settings. Berger’s major contribution lies in advancing depth perception for non-Lambertian and specular surfaces—materials like glass and metal that confound traditional passive and active range sensors. His most cited work, "Depth from stereo polarization in specular scenes for urban robotics" (2017, 34 citations), introduces a novel method that integrates polarization cues with stereo vision to reliably recover 3-D structure from highly reflective objects without artificial illumination. This work is complemented by his earlier paper on incorporating polarization into stereo vision for non-Lambertian scenes (2016, 8 citations), which further establishes his expertise in handling complex reflectance. Beyond perception, Berger has contributed to distributed systems visualization with the vIsage framework (2009), demonstrating versatility. His research directly addresses a critical bottleneck in urban robotics, enabling safer and more robust navigation for autonomous vehicles and robots operating in environments dominated by glass facades and shiny surfaces.
Research Focus
Key Achievements
Top Papers
- 1Depth from stereo polarization in specular scenes for urban robotics34 citations · 2017
- 2
- 3