Curriculum Vitae

Since September 2026, Yuxuan He has been a Ph.D. candidate in Computational Humanities at the Otto-Friedrich University of Bamberg under the supervision of Prof. Dr.-Ing. Jakob Abeßer.

Yuxuan He’s research focuses on machine listening, particularly the development of efficient deep learning models for acoustic scene classification and continual learning. His work aims to enable audio-processing systems to recognize complex acoustic environments with low computational and energy requirements while continuously adapting to new acoustic conditions without losing previously acquired knowledge. His research interests include audio signal processing, resource-efficient machine learning, and intelligent hearing technologies.

Yuxuan He received his bachelor’s degree in psychology, with a minor in Linguistics, from the University of Alberta in Canada in 2017. He subsequently completed the M.Sc. program in Psychology: Learning Sciences at Ludwig-Maximilians-Universität München. In his master’s thesis, he investigated the automatic recognition of teachers’ emotions from vocal cues. His research combined speech segmentation, acoustic feature extraction, and deep-learning methods to develop models for detecting emotional information in speech signals. From 2020 to 2023, he worked as a HCI developer at the intersection of artificial intelligence, psychology, and human-computer interaction. His work focused on developing AI models for vocal emotion recognition and integrating them into user-centered applications. In particular, he applied these models in virtual-reality user studies in the field of architectural design, where participants’ emotional responses to different spatial environments were captured and analyzed. 

In 2023, he joined the interdisciplinary research project “NeuroSensEar – Neuromorphic Acoustic Sensing for the High-Performance Hearing Aids of Tomorrow”. Funded by the Carl Zeiss Foundation, the project develops bio-inspired, adaptive, and energy-efficient technologies for future hearing devices. Within NeuroSensEar, his research focuses on the development of efficient models for acoustic scene classification. He designs and evaluates compact deep-learning architectures and multiscale feature representations that can recognize complex acoustic environments while reducing computational and energy requirements. He also investigates continual-learning methods that enable acoustic models to adapt progressively to new environments and previously unseen listening situations without losing previously acquired knowledge. His doctoral research builds on this work, with a particular emphasis on balancing classification accuracy, computational efficiency, adaptability, and real-time performance for resource-constrained applications such as hearing aids.