报告题目：Exploring Cutting-Edge Techniques for Speech Emotion Recognition in Our Research Group
报告人：张帆（浙江大学百人计划研究员）
时间：2024年6月28日（星期五）上午10:00
腾讯会议：947-897-675
邀请人：李瑾
报告摘要：Speech Emotion Recognition (SER) enhances human-computer interactions by making technology more empathetic. This presentation explores both classical methods and modern deep learning techniques in SER, highlighting their potential applications in physics, particularly in the detection of gravitational waves (GWs) within the audio frequency band.We review classic SER methods like Hidden Markov Models and Gaussian Mixture Models, and then focus on advanced deep learning methods such as self-attention mechanisms and multiscale area attention. These methods are analogous to techniques used in analyzing GW signals, which also require precise detection and classification of subtle patterns within noisy data.Data augmentation techniques such as noise injection and time stretching, used to improve SER model robustness, can similarly enhance the detection and analysis of GW signals. Furthermore, our confidence-based multimodal fusion approach, which integrates speech and text emotion recognition models, demonstrates significant performance improvements that could be applied to multi-modal GW detection systems.
报告人介绍：Fan Zhang has held a series of research and development positions throughout his career. He served as a research scientist at both IBM Massachusetts lab and the Kavli Institute for Astrophysics and Space Research at Massachusetts Institute of Technology, and was a sponsored researcher at Tsinghua University in Beijing, China. Additionally, he has held the position of visiting professor at the Nanjing Tech University.In 2012, he earned his Ph.D. from the Department of Control Science and Engineering at Tsinghua University. He later worked as a research scientist at the Cloud Computing Laboratory at Carnegie Mellon University from 2011 to 2013. He has been recognized with several awards and honors throughout his career, including an Honorarium Research Funding Award from the University of Chicago and Argonne National Laboratory, a Meritorious Service Award from IEEE Transactions on Service Computing, and two IBM Ph.D. Fellowship Awards.
