“Advancements in Spoken Language Models: Building upon Text-Based Large Language Models” Hung Yi Lee
Автор: TTIC
Загружено: 2025-09-25
Просмотров: 105
“Advancements in Spoken Language Models: Building upon Text-Based Large Language Models”
Hung-yi Lee, National Taiwan University
Originally recorded on September 5, 2025 as part of the TTIC Summer Workshop on Foundations of Speech and Audio Foundation Models.
In this talk, Hung-yi Lee presents recent advancements in spoken language models (SLMs), which extend text-based large language models to process and generate speech while preserving their core capabilities. He introduces new speech representation learning techniques, benchmark evaluations, and methods for enabling full-duplex dialogue—pushing forward the integration of speech into the LLM framework.
Timestamps:
00:00 Introduction
01:56 Talk begins
57:36 Q&A
#SpokenLanguageModels #SpeechAI #AI #MachineLearning #ML #NLP #NaturalLanguageProcessing #MultimodalAI #DeepLearning #LLMs #TTIC #languagemodels
Доступные форматы для скачивания:
Скачать видео mp4
-
Информация по загрузке:
