[Tutorial I]: 15:20 ~ 16:10, February 24, 2026 (Tuesday)
Prof. Takumi Takahashi (Osaka University, Japan)
Fundamentals of Bayesian Message-Passing Algorithms for Generalized Linear Models
Abstract:
This tutorial focuses on generalized linear models (GLMs), which frequently arise in physical-layer signal processing for wireless communications, and introduces the fundamentals of probabilistic inference algorithms for solving inverse problems based on GLMs. In particular, we highlight message-passing algorithms (MPAs) grounded in Bayesian inference, which can achieve optimal estimation under certain conditions. Starting from the most basic sum-product algorithm (SPA), we then proceed to Gaussian belief propagation (GaBP), which appears in a wide range of inference problems as a form of large-scale belief propagation, and finally provide a detailed derivation leading to the best-known MPA, generalized approximate message passing (GAMP). We conclude by presenting our recent research contributions related to Bayesian MPAs and their applications to wireless communications.
Biography
Takumi Takahashi (Member, IEEE) received the B.E., M.E., and Ph.D. degrees in communication engineering from The University of Osaka, Osaka, Japan, in 2016, 2017, and 2019, respectively. From 2018 to 2019, he was a Visiting Researcher at the Centre for Wireless Communications, University of Oulu, Finland. In 2019, he joined the Graduate School of Engineering, The University of Osaka, as an Assistant Professor, where he is currently an Associate Professor. His research focuses on signal processing and wireless communications. His current research interests include coding theory, Bayesian inference, compressed sensing, and MIMO technologies.
[Tutorial II]: 10:40 ~ 11:30, February 26, 2025 (Thursday)
Prof. Soonmin Hwang (Hanyang University, Republic of Korea)
Camera-Based Perception for Autonomous Driving: Fundamentals to End-to-End Learning
Abstract:
This tutorial introduces the principles and methodologies of camera-based perception for autonomous driving. Beginning with camera sensor fundamentals and the role of multi-sensor calibration and synchronization, we systematically cover key perception tasks such as object detection, depth estimation, and trajectory prediction, among others. We then discuss recent trends toward end-to-end autonomous driving, where perception is jointly optimized with downstream decision-making modules. The tutorial concludes with an overview of Tesla’s data engine, highlighting its role in enabling scalable, real-world physical AI systems.
Biography
Soonmin Hwang is an Assistant Professor in the Department of Automotive Engineering at Hanyang University. He was previously a Postdoctoral Researcher at the Robotics Institute, Carnegie Mellon University, and a Senior Machine Learning Scientist on the Autopilot team at Tesla, where he worked on practical machine learning models for autonomous driving. He received his Ph.D. in Electrical Engineering from KAIST in 2019 under the supervision of Prof. In So Kweon. His research focuses on computer vision and machine learning for autonomous driving, bridging academic research and real-world industrial applications. He was selected as an Outstanding Reviewer for CVPR 2023 and is a recipient of the Samsung HumanTech Paper Awards (Gold Prize, Honorable mention) and the First Place in the NVIDIA Korea Deep Learning Contest.