|
11月14日深圳大学毕宿志学术报告预告
作者：
发布日期：2022-11-07
浏览次数：
报告主题：Federated Transfer Learning for Domain-Adaptive Indoor Wireless Sensing 报 告 人：毕宿志 报告时间：2022年11月14日 10:00-11:30 报告网址：#腾讯会议：931-515-759 报告摘要： Wireless indoor sensing that leverages the rich scattering environment can provide various context-related sensing functions, such as crowd, posture and respiration monitoring. The key technical problem is that the sensing accuracy is heavily dependent on the environment, where a well-functioning model finetuned to one domain may perform poorly in another unfamiliar one, even under minor change of background setting or transceiver placement. In this work, we propose potential solutions using transfer and federated learning techniques, combined with sensing data processing methods, to achieve domain-adaptive and sample-efficient sensing performance across dissimilar environments.
报告人简介： Suzhi Bi received his bachelor’s degree from Zhejiang University in 2009, PhD degree from The Chinese University of Hong Kong in 2013, and is now an Associate Professor with College of Electronics and Information Engineering, Shenzhen University. His major research interests include wireless resource allocation, mobile computing, and wireless sensing. He received the APB outstanding young researcher award and five IEEE journal and conference best paper awards. He is now an editor of IEEE Transactions on Wireless Communications and IEEE Wireless Communications Letters. |
