ModaVerse: Efficiently Transforming Modalities with LLMs
Paper
•
2401.06395
•
Published
•
3
Audio LLM
Note Multimodal LLM, Using Agent binding to Generation Model for generation
Boosting Large Language Model for Speech Synthesis: An Empirical Study
Paper • 2401.00246 • Published • 7Note Approach A didn't work well so the authors conclude that the speech tokens cannot be treated as a new language
-
An Integration of Pre-Trained Speech and Language Models for End-to-End Speech Recognition
Paper • 2312.03668 • Published • 1 -
Towards General-Purpose Speech Abilities for Large Language Models Using Unpaired Data
Paper • 2311.06753 • Published • 5 -
Qwen-Audio: Advancing Universal Audio Understanding via Unified Large-Scale Audio-Language Models
Paper • 2311.07919 • Published • 7 -
Audiobox: Unified Audio Generation with Natural Language Prompts
Paper • 2312.15821 • Published • 12 -
ZMM-TTS: Zero-shot Multilingual and Multispeaker Speech Synthesis Conditioned on Self-supervised Discrete Speech Representations
Paper • 2312.14398 • Published
SpeechGPT: Empowering Large Language Models with Intrinsic Cross-Modal Conversational Abilities
Paper • 2305.11000 • Published • 3Note Using HuBERT speech token directly in the LLM. Train a GAN vocoder (HiFi-GAN) for decoding.
-
fnlp/SpeechInstruct
Viewer • Updated • 3 • 9 -
SpeechTokenizer: Unified Speech Tokenizer for Speech Large Language Models
Paper • 2308.16692 • Published • 1 -
Audiogen/agc-continuous
Updated • 4.35k • 1 -
Audiogen/agc-discrete
Updated • 6.43k • 2 -
hubertsiuzdak/snac_44khz
Updated • 182 • 2 -
hubertsiuzdak/snac_32khz
Updated • 20.6k • 3 -
hubertsiuzdak/snac_24khz
Updated • 1.85k • 4