Papers
7
Total Citations
165
H-Index
6
About
Xiaoning Han is a leading researcher in intelligent robotics and computer vision, whose work bridges the gap between autonomous perception and real-world robotic action. His primary research areas include active object detection, semantic mapping, and robot learning from demonstration. Han’s most impactful contribution is his pioneering work on active object detection using deep reinforcement learning, where he developed a multistep action prediction framework with Deep Q-Networks (88 citations) and advanced the field with Double DQN and prioritized experience replay (12 citations). These methods enable robots to intelligently adjust their viewpoints to overcome challenges like occlusion and partial capture—critical for practical deployment. He also authored a comprehensive survey on semantic mapping for mobile robots (27 citations), establishing a foundational reference for attaching semantic meaning to geometric maps in indoor scenes. Han’s work on obstacle avoidance learning using mixture models (13 citations) and plane extraction from inverse depth images (12 citations) further demonstrates his commitment to robust, human-inspired robotic navigation. His research on complex sequential task learning with Bayesian inference (8 citations) advances skill transfer from humans to robots. With over 165 total citations, Han’s contributions are shaping the next generation of autonomous systems capable of perceiving, understanding, and navigating dynamic environments.
Research Focus
Key Achievements
Top Papers
- 1
- 2Semantic Mapping for Mobile Robots in Indoor Scenes: A Survey27 citations · 2021
- 3Robot Obstacle Avoidance Learning Based on Mixture Models13 citations · 2016
- 4Active Object Detection Using Double DQN and Prioritized Experience Replay12 citations · 2018
- 5A Plane Extraction Approach in Inverse Depth Images Based on Region-Growing12 citations · 2021
- 6
- 7A novel navigation scheme in dynamic environment using layered costmap5 citations · 2017