SOTA discrete acoustic codec models with 40 tokens per second for audio language modeling

FunCodec is a research-oriented toolkit for audio quantization and downstream applications, such as text-to-speech synthesis, music generation et.al.

unofficial implementation of the High Fidelity Neural Audio Compression

SimVQ: Addressing Representation Collapse in Vector Quantized Models with One Linear Layer

Codebase for the paper 'EncodecMAE: Leveraging neural codecs for universal audio representation learning'

Simplified installers for suno-ai/bark, musicgen, tortoise, RVC, demucs and vocos

通过哈夫曼树编解码原理编写解码器，实现文件的压缩与解压缩

Nvidia video hard decoding, rendering, soft/hard encoding, and writing to MP4 file ; Nvidia视频硬解码、渲染、软/硬编码并写入MP4文件

A TTS app where you can clone the voices of any person you wish.

Experiments sonifying frame-level encodecmae features and encodecmae summary vectors using generative audio models.

A collections of audio codecs with a standardized API

Go binding for encodec.cpp

A Survey of Spoken Dialogue Models (60 pages)