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 audio, music, and speech generation research and development.

Easy-to-use Speech Toolkit including Self-Supervised Learning model, SOTA/Streaming ASR with punctuation, Streaming TTS with text frontend, Speaker Verification System, End-to-End Speech Translation and Keyword Spotting. Won NAACL2022 Best Demo Award.

vits2 backbone with multilingual-bert

Python package for NSF and NSF-HiFi-GAN (unofficial)

A collection of Persian text-to-speech models using implementations and techniques.

A DDSP-based neural voice synthesiser.

A high-quality speech analysis, manipulation and synthesis system

Turn your Android phone into Amateur Radio Codec2/OPUS APRS enabled DV handheld transceiver (Bluetooth/BLE/USB/TCPIP KISS/Sound modem client for DV digital voice communication)

Voice Changer VST Plugin

🐸💬 - a deep learning toolkit for Text-to-Speech, battle-tested in research and production

Vocos: Closing the gap between time-domain and Fourier-based neural vocoders for high-quality audio synthesis

Real-Time High-Fidelity Speech Synthesis without GPU

HiFi-GAN: Generative Adversarial Networks for Efficient and High Fidelity Speech Synthesis

A toolkit for non-parallel voice conversion based on vector-quantized variational autoencoder

VocGAN: A High-Fidelity Real-time Vocoder with a Hierarchically-nested Adversarial Network

VocalTractLab development repo

Implementations of HiFi-GAN, iSTFTNet and MISRNet

Official JUCE plugin for stftPitchShift

A JUCE implementation of the Pitch Vocoder using COLA, pitch tracking (YIN), and handling audio input and Midi Events.

😝 TensorFlowTTS: Real-Time State-of-the-art Speech Synthesis for Tensorflow 2 (supported including English, French, Korean, Chinese, German and Easy to adapt for other languages)