Fangming Guo
Research Assistant
Department of Computer Science
Chongqing University
About Me
I am a research assistant at the School of Computer Science, Chongqing University, specializing in brain-inspired artificial intelligence algorithms. My goal is to design intelligent systems as capable as the human brain, tackling real-world challenges such as robotic environmental interaction, indoor positioning, and medical applications.
My research interests include computational neuroscience, NeuroAI, and perception. Currently, I am exploring neural data representations to uncover the relationship between behavior and neural activity.
News
-
January 2025Our paper "Event-driven Tactile Sensing With Dense Spiking Graph Neural Networks" has been accepted by IEEE Transactions on Instrumentation and Measurement.
-
November 2024Our paper "Spike-BRGNet: Efficient and Accurate Event-based Semantic Segmentation" has been accepted by IEEE Transactions on Circuits and Systems for Video Technology.
-
March 2024Our paper "Accurate and Efficient Floor Localization with Scalable Spiking Graph Neural Networks" has been published in Satellite Navigation.
-
December 2023Presented our work "Efficient and Accurate Indoor/Outdoor Detection with Deep Spiking Neural Networks" at IEEE GLOBECOM 2023 conference.
Featured Publications
Event-driven Tactile Sensing With Dense Spiking Graph Neural Networks
IEEE Transactions on Instrumentation and Measurement, 2025
This work proposes DeepTactile, a spiking graph neural network designed for event-driven tactile sensing. By structuring tactile data as graphs based on the local connectivity of taxels and leveraging spiking graph convolutional networks for feature extraction, the method effectively handles event-based learning.
Spike-BRGNet: Efficient and Accurate Event-based Semantic Segmentation with Boundary Region-guided Spiking Neural Networks
IEEE Transactions on Circuits and Systems for Video Technology, Nov 2024
This work proposes Spike-BRGNet, a spike-driven boundary region-guided network for event-based semantic segmentation in traffic scenes. By leveraging a three-branch spiking encoder, a spiking multi-scale context aggregation module, and a novel boundary region-guided loss function, the model effectively captures spatial-temporal dynamics from event streams.