Title: Better audio through AI research at Bose

Shuo Zhang, Principal Machine Learning Research Engineer at Bose Research 

Abstract: 

In an era of LLMs and large audio models, how do we translate the power of SOTA deep learning and AI into great audio experiences on embedded devices? At Bose Research, we take on the challenge of using AI and ML to make our audio products more immersive, intelligent and contextual. In this talk, I will focus on the role of deep learning and AI research that powers our next generation on-device audio experiences. We will highlight current key challenges and opportunities in prototyping SOTA deep learning algorithms on compute constrained platforms for real-time audio experiences. Finally, I will give an overview of several application areas and research outputs, including audio understanding, source separation, generative audio, and tinyML.

Bio:

Shuo Zhang is a machine learning researcher and engineer at Bose Research, where he works on research and development of NLP and audio applications. Previously he was a Collaborator and Researcher at the Music Technology Group (MTG), Universitat Pompeu Fabra, Barcelona, Spain. He received his Ph.D from Georgetown University[GUCL Group, CorpLing Lab] with a focus on Computational Linguistics. His research interest includes machine learning and deep learning for audio (speech, music, environmental audio understanding), natural language processing (NLP), speech prosody, and computational musicology. Web: https://zangsir.github.io/

This activity is part of the Cátedra UPF-BMAT en Inteligencia Articial y Música (TSI-100929-2023-1). Project funded by Secretaría de Estado de Digitalización e Inteligencia Artificial, and Unión Europea-Next Generation EU

​