Jiatong Shi

A new method supports the preservation of at-risk languages through detailed documentation.

A new framework for assessing foundation models in speech tasks.

A new model improves speech recognition using multiple decoding methods.

An easy-to-use tool for fine-tuning speech models without complex code.

Discover how preference alignment improves text-to-speech systems for better user experiences.

ESPnet-Codec enhances training and evaluation of neural codecs for audio and speech.

A guide to effectively learning a new language with practical tips.

New methods improve how machines recognize spoken language.

VERSA evaluates speech, audio, and music quality effectively.