1 paper · 2 filters
Dongchao Yang, Yuanyuan Wang, Dading Chong +3
We study two foundational problems in audio language models: (1) how to design an audio tokenizer that can serve as an intermediate representation for both understanding and genera…