IEEE SMC Guangzhou Chapter 2025 Academic Seminar
The IEEE SMC Guangzhou chapter 2025 academic Seminar will be held in Shenzhen on Friday, November 7, 2025, from 13:30 to 15:30.
Organized by the IEEE SMC Guangzhou Chapter, this seminar primarily focuses on frontier artificial intelligence directions. Serving as a key platform for presenting the latest research achievements in related fields, the event is dedicated to establishing a high-level academic exchange platform, promoting the deepening and expansion of cross-disciplinary collaboration, and facilitating in-depth discussions on new development pathways and core challenges of artificial intelligence in the new era, thereby injecting new momentum into the field.
Date and Time
Location
Hosts
Registration
Add Event to Calendar
Speakers
Shanshan Wang
Exploration of Basic Models for Artificial Intelligence in Medical Imaging: A Holistic Perspective from Imaging to Quant
The development of medical multimodal artificial intelligence (AI) is currently grappling with several core challenges, including high data heterogeneity, a scarcity of annotations, and inadequate cross-scenario generalization capabilities. While existing approaches have achieved certain progress in multimodal data integration, in real-world clinical settings, problems such as insufficient modality synergy and limited model adaptability still persist. This report focuses on a "imaging-quantification-analysis" full-chain perspective to explore the construction and application of basic models for medical imaging artificial intelligence. We propose a pre-training strategy that integrates self-supervised and contrastive learning to fully exploit the cross-modal correlations in large-scale unlabeled medical data, thereby reducing the reliance on manual annotations and enhancing the robustness and generalization ability of the model in different institutions, devices, and population environments. Based on this research idea, the team has achieved initial results in several key application scenarios, including efficient and robust fast magnetic resonance imaging, multimodal lung imaging and clinical text understanding, as well as multimodal multi-scale alignment and zero-shot adaptive learning for open environments. Finally, we will present the team's recent work published in prestigious journals such as Nature Biomedical Engineering, Scientific Data, and Nature Communications over the past year. Thess works cover the construction of medical imaging basic models, neuroimaging microstructure imaging, and clinical disease applications, demonstrate the full-chain exploration and practice of medical imaging AI from imaging to analysis.
Biography:
Wang Shanshan is a researcher and Ph.D. supervisor at the Shenzhen Institute of Advanced Technology, Chinese Academy of Sciences. She is a recipient of the National Science Fund for Excellent Young Scholars and the Wu Wen Jun AI Outstanding Youth Award, an outstanding Class A member of the Youth Innovation Promotion Association of CAS, and has been selected for the Elsevier and Stanford "World's Top 2% Scientists" list from 2021 to 2025. She has long been engaged in research in artificial intelligence, fast medical imaging, radiomics, and multimodal analysis, having published over 100 high-quality papers in journals such as Nature BMEand IEEE Trans. She has done pioneering work in AI-based fast medical imaging and imaging physics-guided small-sample learning techniques. She was invited to deliver a named plenary lecture at the 31st Annual Meeting of the International Society for Magnetic Resonance in Medicine (acceptance rate approximately 1/6000; London, UK) and a plenary lecture at the 10th GRC In Vivo Magnetic Resonance Conference (Andover, USA). Core technologies developed by her have been transferred to two medical companies and installed in over 1,000 units, including various MRI and low-dose CT devices. She has received awards including the Wu Wen Jun AI Technology Invention First Prize (Rank 1/6), the OCSMRM Outstanding Research Award (Rank 1/1), the Guangdong Technology Invention and Technological Progress First Prize (Rank 2/10), and the Guangdong Youth Science and Technology Award (Rank 1/1). She has led six national-level projects, including the MoST 2030 "New Generation AI" Major Project, the NSFC Joint Fund Key Project, and the Excellent Young Scholars Project. She serves as Associate Editor or editorial board member for several high-quality SCI journals, such as IEEE Transactions on Medical Imaging, IEEE Transactions on Computational Imaging, Magnetic Resonance in Medicine, Pattern Recognition, and Biomedical Signal Processing and Control
Daojing He
Some Practices on How Artificial Intelligence Can Help Industries Grow Stronger and Larger
Introduce the team's practical application of artificial intelligence and robot technology in several typical fields.
Biography:
Professor Daojing He is a tenured faculty member at Harbin Institute of Technology (HIT), Shenzhen. He currently serves as Vice Dean of the School of Computer Science and Executive Vice Dean of the Institute of Computing and Intelligence. He also directs the Guangdong Provincial Engineering Research Center for Cryptographic Application Innovation in Higher Education Institutions.
Baoyuan Wu
The Evolution of Deepfake Detection: From Generalization and Interpretability to Continuous Adaptation
In this talk, we’ll take you on a journey through the development of deepfake generation and detection technologies, introducing for the first time a “generational evolution” framework to help you systematically understand the core characteristics of each era. We’ll analyze breakthroughs and limitations through three critical lenses: generalization, interpretability, and evolutionary adaptability. We will then walk through the representative detection techniques of each generation, highlighting their key innovations and practical impact. Building on this foundation, we will offer a forward-looking perspective on the defining features and developmental trajectory of next-generation detection systems — particularly those powered by multimodal large models.
Biography:
Dr. Baoyuan Wu is a Tenured Associate Professor of School of Artificial Intelligence, The Chinese University of Hong Kong, Shenzhen, Guangdong, 518172, P.R. China. His research interests are trustworthy and generative AI, as well as optimization. He has published 100+ top-tier conference and journal papers, including 13 TPAMI and IJCV, and one paper was selected as the Best Paper Finalist of CVPR 2019. His Google Scholar citation is 11,000+. He is currently serving as a Senior Area Editor of IEEE TIFS, Associate Editors of Neurocomputing and Frontiers in Computer Science (Computer Vision Section), Organizing Chair of PRCV 2022, Area Chairs of CVPR 2024/2025/2026、NeurIPS 2022/2023/2024/2025、NeurIPS Datasets and Benchmarks Track 2023/2024/2025, ICLR 2022/2023/2024/2025/2026, ICML 2023/2024/2025, AAAI 2022/2024/2025 /2026, and AISTATS 2024. He is IEEE Senior Member. He was selected into the world's top 2% scientist list of 2021~2024, released by Stanford University. He was awarded the “2023 Young Researcher Award” of The Chinese University of Hong Kong, Shenzhen.