MATS: An Audio Language Model under Text-only Supervision.

Wen Wang, Ruibing Hou, Hong Chang, Shiguang Shan, Xilin Chen

Browse the full ICML paper archive.