Biography: Yuan Wu is currently the Associate Professor with the State Key Laboratory of Internet of Things for Smart City, University of Macau, Macao, China, and also with the Department of Computer and Information Science, University of Macau. His research interests focus on edge computing and edge intelligence. He received the Best Paper Award from IEEE ICC’2016, IEEE TCGCC’2017, and WCNC’2023. He served as the symposium TPC co-chairs for GLOBECOM’2024, ICCC’2023, VTC’2022-Spring, etc. He is currently on the editorial board of IEEE TWC, TVT, and TNSE. He is a senior member of IEEE and distinguished member of CCF, and he currently serves as the Vice Chair of IEEE ComSoC Macau Chapter.
Talk Title:
Revolutionizing Mobile Edge Networks with Generative AI: From Efficient Architectures to Emerging Applications
Talk Abstract :
The integration of generative AI into mobile edge networks has the potential to transform our daily lives and interactions. This talk will explore two critical questions: how to efficiently train, finetune and deploy large models in edge networks, and how generative AI models can transform edge network applications. We will explore recent breakthroughs in efficient personalized fine-tuning techniques, including low-rank adaptation (LoRA) and reinforcement learning from human feedback (RLFH), which enable efficient adaptation and alignment of large models. Our cost-adjustable framework for generative AI services in edge computing will also be introduced, providing flexible and efficient deployment options. Furthermore, we will delve into cutting-edge applications that have the potential to revolutionize industries. For instance, we will discuss how generative AI can be leveraged to rectify data imbalances in federated learning, thereby enhancing fairness and accuracy. Furthermore, we will explore the deployment of large multi-modal models for autonomous driving, which can significantly enhance safety and efficiency on our roads. By harnessing the power of generative AI, we can unlock new possibilities for mobile edge networks and create a more intelligent, connected, and automated future.
