Autoregressive Guidance of Deep Spatially Selective Filters using Bayesian Tracking for Efficient Extraction of Moving Speakers
This episode discusses a paper focusing on extracting moving speakers using Bayesian tracking combined with autoregressive guidance. The authors address dynamic sound sources by creating a robust system that manages uncertainty about a speaker's path. This method uses feedback from the enhanced speech to improve accuracy while remaining computationally efficient for practical, real-time use.