Welcome 

Hongyu's group focuses on exploring and developing neuromorphic computing and its applications. Our research involves several aspects, including AI hardware, novel brain-inspired learning algorithms, and their applications in robotics, autonomous systems, and medical devices.To advance these scientific fields, we dedicate our efforts to three main research directions: 

We operate Neuromorphic Robotics Lab and Neuromorphic Brain-Machine Interface Lab. For more information about our latest outcomes, please see publication and research. 

Research Highlight

Demonstrating Self-Learning Robot in Open-Field Arena

Associative_Learning.mp4
Experiment_video_for_slide.mp4

This work replicates the associative learning observed in rodents in the open-field arena by memorizing the causal relationship between conditional and unconditional stimuli.

The self-learning robot automatically memorizes (without pre-training) the relationship between a red-colored wall and the road bumpers when these two stimuli are detected at same time. In this work acceleration (vibration) from the road bumper is assigned as the unconditional and aversive stimulus, while the red-colored wall is assigned as the neutral stimulus. Before associative learning, the neuromorphic robot has no response to red colored wall. After associative learning, the neuromorphic robot evokes an avoidance action (retreating or turning to other directions) when it detects the red color with its camera. 

Related paper: 

Tianze Liu, Md Abu Bakr Siddique, Hongyu An, "Mimicking Associative Learning of Rats via a Neuromorphic Robot in Open Field Maze Using Spatial Cell Models," 2024 International Conference on Neuromorphic Systems (ICONS), Arlington, VA, USA, 2024, pp. 299-306, doi: 10.1109/ICONS62911.2024.00052.

Media Highlight


To Join Our Team

For postdocs, graduate students, and undergraduate students interested in working in our group, please send your CV to hongyua@mtu.edu.