Sebastian Klose

Information Technology University, Technical University of Munich

Papers

Total Citations

H-Index

About

Sebastian Klose is a leading researcher in computer vision and robotics, specializing in real-time dense 3D mapping and visual odometry. His work centers on developing efficient algorithms for motion estimation and depth perception using RGB-D sensors, with a focus on robust performance in dynamic environments. Klose’s most impactful contribution is his 2013 paper on efficient compositional approaches for real-time robust direct visual odometry from RGB-D data (49 citations), which systematically evaluated photometric error minimization techniques for frame-to-frame motion estimation. This work demonstrated how direct image alignment methods could achieve high accuracy and robustness, establishing a foundation for subsequent advances in visual SLAM systems. In his 2015 paper on fast dense stereo correspondences by binary locality sensitive hashing (18 citations), Klose addressed the computational challenges of high-resolution stereo matching, proposing a hashing-based approach that significantly reduces runtime complexity and memory requirements—a practical breakthrough for robotic applications requiring real-time performance. His research bridges the gap between theoretical optimization and deployable systems, making him a notable figure in the field of real-time 3D perception.