Ziwen Dou

Papers

Total Citations

183

H-Index

About

Ziwen Dou is a leading researcher in multi-sensor fusion and self-supervised depth estimation for autonomous systems. His work centers on developing efficient, learning-based calibration and perception methods that eliminate the need for labor-intensive manual setups. Dou’s most influential contribution is LCCNet (2021, 175 citations), a pioneering LiDAR-camera self-calibration framework that leverages a cost volume network to achieve robust, automatic sensor alignment—a critical precondition for reliable 3D reconstruction and environment perception in self-driving and robotics. More recently, Dou introduced a lightweight approach to self-supervised monocular depth estimation (2024), addressing the growing demand for computationally efficient models in real-world deployment. By proposing a deep neighbor layer aggregation strategy, his method achieves competitive depth accuracy without relying on large, complex networks, making it highly suitable for resource-constrained platforms. With over 180 total citations and a focus on practical, deployable solutions, Dou’s research bridges the gap between algorithmic sophistication and real-world efficiency, offering tangible advances for autonomous navigation and robotic perception.