Kakaru

ColorSplitter. A cli tool for split vocal timbre.

R3MOE. [RecurrentNN × Regression × Regularized]-base Mouth Opening Estimation via SSL(Semi-supervised Learning).

CODEY_Dataset. 一个第三方的泠鸢yousa歌声数据集

ds2svc. Convert the "DiffSinger" model to an SVC (such as RVC) model

Asaritsu4Diffsinger. A mult-languages(CN/JP) singing database for Diffsinger(OpenVPI).

GPT-SoVITS-musa. 1 min voice data can also be used to train a good TTS model! (few shot voice cloning)

DiffSinger. An advanced singing voice synthesis system with high fidelity, expressiveness, controllability and flexibility based on DiffSinger: Singing Voice Synthesis via Shallow Diffusion Mechanism

rectified-flow. 从零手搓Flow Matching（Rectified Flow）

LaBei4DiffSinger. UTAU音库LaBei Rasubeku的DiffSinger移植

auorange-torch. Audio LPC (linear prediction code) using mel spectorgram, compatible for LPCNet, torch version

yousa-ling-diffsinger-v1. 泠鸢yousa的Diffsinger模型v1版

libmininsf. C implementation of the MiniNSF source generator used by OpenVPI DiffSinger's NSF-HiFiGAN vocoder.

game_ggml_cli. C++

audio-preprocessing-scripts. 数据集自动化制作脚本

mel2lpc_torch. Using Mel spectrum to calculate linear predictive coding coefficients (LPC), implemented by pytorch

onedrive-vercel-index. OneDrive public directory listing, powered by Vercel and Next.js

dataset-tools-mod4ass. A mod for dataset-tools(OpenVPI) for support .ass file(Advanced SubStation Alpha).

ROSA-FPGA. A native ROSA (Rapid Online Suffix Automaton) implementation on Cyclone IV EP4CE10F17.