Kohei Watanabe

Papers

Total Citations

184

H-Index

About

Kohei Watanabe is a leading researcher in autonomous vehicle perception and computer vision, with a primary focus on multispectral object detection and domain-adaptive semantic segmentation. His most influential work, "Multispectral Object Detection for Autonomous Vehicles" (2017), has garnered 177 citations and addresses the critical challenge of robustly detecting diverse objects—cars, pedestrians, and bicycles—under varying environmental conditions, a cornerstone for safe automated driving. Watanabe’s contributions extend to advancing multichannel semantic segmentation through unsupervised domain adaptation, enabling models to generalize across different sensor modalities and real-world scenarios without labeled target data. His research bridges the gap between theoretical computer vision and practical deployment in mobile robotics, emphasizing reliability in complex traffic environments. By integrating multispectral data (e.g., thermal and visible light), Watanabe enhances detection accuracy in low-visibility settings, a notable achievement for autonomous systems. His work is highly cited and foundational for researchers developing resilient perception pipelines, making him a key figure in the evolution of autonomous vehicle technology.