University of Southern California（南加州大学）

Reconstruct! Don't Encode: Self-Supervised Representation Reconstruction Loss for High-Intelligibility and Low-Latency Streaming Neural Audio Codec

重建！不要编码：面向高可懂性和低延迟流式神经音频编解码器的自监督表示重建损失

Junhyeok Lee, Xiluo He, Jihwan Lee, Helin Wang, Shrikanth Narayanan, Thomas Thebaud, Laureano Moro-Velazquez, Jesús Villalba, Najim Dehak

机构 * Center for Language and Speech Processing, Johns Hopkins University, USA（语言与语音处理中心，约翰霍普金斯大学，美国） ； Signal Analysis and Interpretation Laboratory, University of Southern California, USA（信号分析与解释实验室，南加州大学，美国） 纠错

AI总结 本文提出自监督表示重建损失，用于提升流式神经音频编解码器的可懂性和低延迟性能。

Comments Submitted to Interspeech 2026