Graph learning for multi-agent perception and prediction

May 13, 2022, zoom / ESB 2001

Siheng Chen

Shanghai Jiao Tong University, Engineering


From a school of fish to a biological neural network with billions of neurons, the nature shows the power of connections. What would it take to allow machines to understand complex, implicit social connections in human society, and even build their own connections? In this talk, we introduce graph learning as a general approach to model and uncover implicit connections. Based on the emerging unrolling techniques, we consider a graph learning framework that leverages both mathematical designs and end-to-end learning ability. We further talk about its applications to autonomous systems. For multi-agent perception, we learn a communication graph that can coordinate multiple agents to strategically collaborate with each other and better perceive a shared scene. For multi-agent prediction, we learn an interaction graph that captures the underlying social connections of multiple agents, promoting more precise and interpretable behavior prediction.

Speaker's Bio

Siheng Chen is a tenure-track associate professor of Shanghai Jiao Tong University. Before joining Shanghai Jiao Tong University, he was a research scientist at Mitsubishi Electric Research Laboratories (MERL), and an autonomy engineer at Uber Advanced Technologies Group (ATG), working on the perception and prediction systems of self-driving cars. Before joining industry, Dr. Chen was a postdoctoral research associate at Carnegie Mellon University. Dr. Chen received his doctorate in Electrical and Computer Engineering from Carnegie Mellon University in 2016, where he also received two master degrees in Electrical and Computer Engineering (College of Engineering) and Machine Learning (School of Computer Science), respectively. He received his bachelor’s degree in Electronics Engineering in 2011 from Beijing Institute of Technology, China. Dr. Chen's work on sampling theory of graph data received the 2018 IEEE Signal Processing Society Young Author Best Paper Award. His co-authored paper on structural health monitoring received ASME SHM/NDE 2020 Best Journal Paper Runner-Up Award and another paper on 3D point cloud processing received the Best Student Paper Award at 2018 IEEE Global Conference on Signal and Information Processing. Dr. Chen contributed to the project of scene-aware interaction, winning MERL President's Award. His research interests include graph signal processing, graph neural networks and group intelligence.

Video URL: