[Under development] SpeechJudge: Towards Human-Level Judgment for Speech Naturalness (https://speechjudge.github.io/)

DiFlow-TTS: Discrete Flow Matching with Factorized Speech Tokens for Low-Latency Zero-Shot Text-to-Speech

Official implementation of "Continuous Autoregressive Language Models"

An official implementation of Flow-GRPO: Training Flow Matching Models via Online RL

T5Voice is a lightweight PyTorch implementation of T5-based text-to-speech synthesis, supporting both streaming and non-streaming speech synthesis with zero-shot capabilities.

Official repository of the IEEE SLT 2024 paper "Self-Supervised Syllable Discovery Based on Speaker-Disentangled HuBERT"

Implementation of "Look, Listen and Recognise:character-aware audio-visual subtitling"

UniVoice: Unifying Autoregressive ASR and Flow-Matching based TTS with Large Language Models

An Open-Source Project to Unify Audio Processing and Generation

SoulX-Podcast is an inference codebase by the Soul AI team for generating high-fidelity podcasts from text.

Make text LLMs listen and speak

Inference code for Interspeech 2025 paper, "LSCodec: Low-Bitrate and Speaker-Decoupled Discrete Speech Codec"

Official repo for "MVDR Beamforming for Cyclostationary Processes".

Text-Only Domain Adaptation for Pretrained Speech Recognition Transformers

Real-time streaming voice anonymization & voice conversion

a simple transcription library for rust

Official Implementation of "Conan: A Chunkwise Online Network for Zero-Shot Adaptive Voice Conversion"

Makepad is a creative software development platform for Rust that compiles to wasm/webGL, osx/metal, windows/dx11 linux/opengl

Tiny data-over-sound library

Rust library for Weighted Finite States Transducers as decribed by Mohri and Allauzen

Speech Resynthesis using Conditional Flow Matching and HuBERT Units

Tiny, no-nonsense, self contained, Tensorflow and ONNX inference

Amphion (/æmˈfaɪən/) is a toolkit for Audio, Music, and Speech Generation. Its purpose is to support reproducible research and help junior researchers and engineers get started in the field of audi…

Fine tune Gemma 3 on an object detection task

Fast and High-Quality Zero-Shot Text-to-Speech with Flow Matching

ONNX neural network inference engine

This is the official repository of ``Scalable Neural Vocoder from Range-Null Space Decomposition'', which is submitted to TPAMI.