Recent Progress in Acoustic Speaker and Language Recognition - October 2013

Explore recent advancements in acoustic speaker and language recognition in this comprehensive lecture by Alan McCree from the Center for Language & Speech Processing at Johns Hopkins University. Gain insights into modern GMM subspace methods, particularly i-vectors, and learn about pattern classification approaches using these features. Discover the effectiveness of simple Gaussian probabilistic models in language recognition and how discriminative training can enhance performance while providing meaningful probability outputs. Delve into Bayesian methods for speaker recognition, including the popular PLDA approach, and understand their success in addressing limited enrollment data per speaker. Examine recent achievements in adapting Gaussian parameters to new domains without labeled training data. Benefit from McCree's extensive experience in speech and signal processing, spanning applications in international speech coding standards, digital answering machines, talking toys, and cellular telephones.