[Seminar] "Towards Extending the Boundary of Non-invasive Brain-Computer Interfaces" by Yansen Wang, Microsoft Research Asia

With the advancement of artificial intelligence and the emergence of EEG-based foundation models, non-invasive EEG-based brain-computer interfaces (BCIs) have increasingly captured the attention of both academia and industry in recent years. However, due to the inherently low signal-to-noise ratio, high dynamism, and individual variability of EEG signals, designing BCIs that are truly applicable to daily life remains a major open challenge in the research community.

 

In this talk, we will briefly outline the key challenges in designing non-invasive BCIs and present some of our latest research efforts from two perspectives: perception and control. On the perception side, we propose a novel EEG data collection paradigm that aligns brain signals with visual and motor experiences to help us better understand how the human brain perceives the world. On the control side, we explore the use of decodable imagination-driven data to develop a new paradigm for high-accuracy BCI control. At the end of the talk, we call for greater interdisciplinary collaboration in this field to collectively shape the future of brain-computer interface technology.

Yansen Wang is a Senior Researcher at the Microsoft Research Asia (MSRA) Shanghai AI/ML Group. He received his bachelor's degree from Tsinghua University in 2019 and his master's degree in Language Technologies from Carnegie Mellon University (CMU) in 2021. His current research focuses on the intersection of artificial intelligence and neuroscience, including using AI to understand brain mechanisms and leveraging neuroscience discoveries to create more powerful AI systems. He has published several papers of EEG signal processing, Brain-computer Interface, Neuro-inspired AI in top conferences including NeurIPS, ICML, ICLR, etc..