Auto-EQ (DSP + ML)

C++ DSP core with Python ML stubs and a pybind11 bridge. Target platform: Linux (works on Windows for development via CMake).

This repository was re-created after being mistakenly pushed to the wrong GitHub account.
As a result, the full commit history from the original repository is not reflected here.

Build

• Windows PowerShell:

./scripts/build_windows.ps1

• Linux/WSL:

bash scripts/build_linux.sh

Example

After building the Python module, process a WAV with fixed gains:

python examples/process_wav.py input.wav output.wav --gains 6 0 3
``)

Or run ML-driven inference (heuristic fallback if no model):

```bash
python -m pip install -r ml/requirements.txt
python ml/infer.py input.wav output.wav

Layout

• src/, include/: DSP code (biquad + 3-band EQ)
• bindings/: pybind11 module exposing Eq3Band
• tests/: Catch2-based tests
• ml/: feature extraction, toy model, training and inference scripts
• examples/: small Python example