报告人：Prof.Song Guo,CAE Fellow,AAIA Fellow,IEEE Fellow,Member of Academia Europaea,Department of Computer Science and Engineering,Hong Kong University of Science and Technology (香港科技大学)
报告时间：2024年06月06日 (星期四) 10:00-11:00 am
报告地点：腾讯在线会议 977-524-608
报告摘要: Foundation Models like GPT, LLaMA, and DALL-E have been transformative in AI, demonstrating remarkable versatility across tasks. Yet,the full potential of edge computing—with its inherent benefits in cost, latency,and privacy—remains untapped for deploying these models.In this talk,we unveil the concept of Edge-native Foundation Models, an innovative approach that harnesses the power of distributed, diverse,and collaborative edge computing resources.We introduce a user-friendly Edge FM-as-a-service system, allowing seamless access to Foundation Model services without the burdens of expensive deployment or intricate management.Furthermore,we present a novel,environment-responsive adaptation strategy for Edge-native FMs,enabling rapid tuning to meet the dynamic demands of edge environments.Crucially,our methodology emphasizes a commitment to ethical standards and regulatory compliance for AI governance at the edge.We envisage a future where Foundation Models are conceived, nurtured,and utilized within the edge ecosystem.Our empirical findings suggest that Edge-native Foundation Models could level the AI playing field,disrupt the centralization of data processing,and offer a viable,scalable architecture for AI's evolution.
基础模型如GPT、LLaMA和DALL-E在人工智能领域发挥了重大作用，展示了在各种任务中的出色通用性。然而，边缘计算的全部潜力——包括成本、延迟和隐私方面的固有优势——尚未被充分利用来部署这些模型。在本次演讲中，我们揭示了边缘原生基础模型的概念，这是一种创新的方法，利用分布式、多样化和协作的边缘计算资源的力量。我们介绍了一个用户友好的边缘基础模型即服务系统，允许无缝访问基础模型服务，而无需进行昂贵的部署或复杂的管理。此外，我们提出了一种新颖的、响应环境的边缘原生基础模型的适应策略，实现了快速调整以满足边缘环境的动态需求。关键是，我们的方法强调了对道德标准和法规合规性的承诺，以实现边缘人工智能治理。我们展望未来，基础模型将在边缘生态系统中构思、培育和利用。我们的实证研究结果表明，边缘原生基础模型可以平衡人工智能竞争环境，打破数据处理的集中化，并为人工智能的演进提供一种可行、可扩展的架构。
报告人简介:Song Guo is a full professor in the Department of Computer Science and Engineering at Hong Kong University of Science and Technology. Prof. Guo made fundamental and pioneering contributions to the development of edge AI and cloud-edge computing which has created significant impact from generation of new scientific knowledge to creation of innovative technologies. He published many papers in top venues and received over a dozen Best Paper Awards from IEEE/ACM conferences, journals and technical committees. He is the recipient of 2024 Edward J. McCluskey Technical Achievement Award, Gold Medal in 2023 Geneva Inventions Expo, Gold Award in 2023 AsiaWorld-Expo, and Intellectual Property Ambassador Award in 2020 Hong Kong Social Enterprise Competition. Prof. Guo is a Fellow of the Canadian Academy of Engineering,Member of Academia Europaea,and Fellow of the IEEE.Prof. Guo has served on IEEE Fellow Evaluation Committee for both ComSoc and Computer Society.He is the founding and current Editor-in-Chief of IEEE Open Journal of the Computer Society and a member of Steering Committee of IEEE TCC.Prof. Guo has been named on editorial board of a number of prestigious international journals like IEEE TC, IEEE TPDS, IEEE TCC, etc. He has also served as chair of organizing and technical committees of numerous IEEE/ACM conferences and workshops. He has served on RGC engineering panel and been frequently invited for various national and international grant/award reviews.
Song Guo是香港科技大学计算机科学与工程系的全职教授。郭教授在边缘人工智能和云边计算的发展中作出了基础性和开创性的贡献，从而产生了重大影响，包括从新科学知识的产生到创新技术的创建。他在顶级期刊和会议上发表了许多论文，并获得了IEEE/ACM会议、期刊和技术委员会颁发的十多个最佳论文奖。他是2024年Edward J. McCluskey技术成就奖的获得者，2023年日内瓦发明博览会金奖获得者，2023年亚洲世界博览会金奖获得者，以及2020年香港社会企业大赛知识产权大使奖的获得者。郭教授是加拿大工程院院士、欧洲科学院院士和IEEE会士。他曾在IEEE Fellow评审委员会担任通信学会和计算机学会的评审委员。他是IEEE计算机学会开放期刊的创始人和现任主编，并担任IEEE TCC指导委员会成员。郭教授入选了多个著名国际期刊的编委会，如IEEE TC、IEEE TPDS、IEEE TCC等。他还曾担任过多个IEEE/ACM会议和研讨会的组织和技术委员会主席。他曾担任香港研究资助局工程评审小组成员，并经常受邀参加各种国家和国际资助/奖项评审。
邀请人：蒋洪波
联系人：胡靖阳 联系电话：13667383125
