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

Composable transformations of Python+NumPy programs: differentiate, vectorize, JIT to GPU/TPU, and more

Convert images of LaTex math equations into LaTex code.

Contrastive Language-Audio Pretraining

Official implementation of "Separate Anything You Describe"

dawproject-py – A Python repository with code for parsing, generating, and modifying DAWProject files, enabling seamless DAW interoperability.

Convert .SFZ files into Bitwig Studio multisample instruments.

Companion code of DAFx23 "Differentiable Feedback Delay Network for Colorless Reverberation"

Spatial Audio Metrics (SAM) is a toolbox to analyse spatial audio and spatial audio perceptual experiments

Wwise-MCP is a Model Context Protocol server that allows LLMs to interact with the Wwise Authoring application. It exposes tools from a custom python waapi function library to MCP clients.

Generates mkdocs files from C++ source code with minimal configuration. Python + libclang, does not use doxygen. Works in docker

Unofficial Conan 2.0 recipe for JUCE Framework

基于MCP协议的AI研究论文生成工具 - AstroInsight Research Assistant。专为科研工作者设计，提供智能论文检索、事实信息提取、研究假设生成等完整研究流程。

waapi封装