A differentiable version of SPTK

In this project we combine techniques from neural voice cloning and musical instrument synthesis to achieve good results from as little as 16 seconds of target data.

GRAFX: An Open-Source Library for Audio Processing Graphs in PyTorch

A DDSP-based neural voice synthesiser.

MIDI Piano synthesizer using DDSP.

Official Repository of Paper: "Towards High-Quality Zero-Shot Singing Voice Conversion in Low-Resource Scenarios"(AAAI 2026)

FLAMO: Frequency-sampling Library for Audio-Module Optimization

Fast and differentiable time domain all-pole filter in PyTorch.

Multitrack music mixing style transfer given a reference song using differentiable mixing console.

Differentiable dynamic range controller in PyTorch.

Searching for Music Mixing Graphs: A Pruning Approach

Python scripts for AI voice changers

HW+SW project aiming to provide guitarists with audio effects generated from a music file

This project aims at filtering out the human vocals of songs using a library Spleeter, and turning them into instrumental versions, creating an ideal platform for young talented music enthusiasts to explore and learn to play various instruments. The vocals are converted to covers using the Differentiable Digital Signal Processing (DDSP) library, which enables direct integration of signal processing elements with deep learning techniques.

Streamlit version for the DDSP timbre transfer demo.

Master's Thesis Project - NYU Steinhardt Music Technology - The raivBox is a standalone neural audio synthesis device based on Google Magenta's DDSP, with an NVIDIA Jetson Nano 2GB embedded development board at its core.

Incomplete DDSP implementation in Pytorch

ddsp-vst ported to AAP (Audio Plugins For Android) using aap-juce

Using DDSP to learn and synthesize singing with modifiable latent space parameters

react-tonejs-crossfader-mixer with upcoming ddsp-ai implementation

reference implementation and experiment scripts that accompany our WASPAA 2025 submission Differentiable Karplus‑Strong Synthesis for Neural Resonance Optimisation.

Knowledge Distillation of different DDSP Decoders (GRU, TCN, S4)

digitally process audio data with AI/ML & Magenta´s DDSP

IOS application developed using Machine Learning models basic-pitch, DDSP to transform recorded audio to the piano sound while giving corresponding music notes.