Yunda Liu

Papers

Total Citations

H-Index

About

Yunda Liu is a robotics researcher whose work bridges the gap between computer vision, embedded systems, and humanoid locomotion. His key research areas include deep learning-based object detection on resource-constrained platforms, mobile robot localization using LIDAR sensors, and reinforcement learning for bipedal walking control. Liu’s most impactful contribution is his 2018 study on implementing CNN-based object detectors on ARM embedded platforms (16 citations), which demonstrated how deep convolutional neural networks can be deployed for real-time object detection in autonomous robots and vehicles—a critical step toward practical edge AI. He also developed a global localization system for mobile robots using LIDAR (4 citations), enabling efficient self-localization in complex environments. In humanoid robotics, Liu pioneered multi-parameter optimization for stair climbing and introduced a model-free reinforcement learning method combining Q-learning with Radial Basis Function Networks for gait control. His work addresses fundamental challenges in making robots autonomous, perceptive, and physically capable in real-world settings.