Hongkai Wen

University of Warwick, University of Oxford

Papers

Total Citations

258

H-Index

About

Hongkai Wen is a leading researcher at the intersection of robotics, computer vision, and deep learning, with a primary focus on enabling robust and efficient spatial intelligence for autonomous systems. His most influential work, "End-to-end, sequence-to-sequence probabilistic visual odometry through deep neural networks" (244 citations), redefined visual odometry by replacing traditional geometric pipelines with a deep learning approach, demonstrating that neural networks can learn to estimate camera motion directly from raw image sequences. This seminal contribution opened new pathways for data-driven localization in challenging environments. Wen has also made significant strides in practical deployment, tackling the computational bottleneck of 6-DoF visual localization by fusing multi-modal sensory data (9 citations) to achieve real-time performance on resource-constrained platforms. Beyond pure navigation, his research extends to human-robot interaction, where he developed probabilistic frameworks that combine user activity data from wearables with mobile robot sensing for semantic mapping and user localization in domestic settings (5 citations). Through these contributions, Wen has established himself as a key figure advancing the reliability and applicability of vision-based systems for real-world robotics.