inv_cmne.cpp File Reference

Implementation of the CMNE solver (dSPM kernel, z-score rectification, LSTM inference, optional training driver). More...

#include "inv_cmne.h"
#include <Eigen/Eigenvalues>
#include <QDebug>
#include <QCoreApplication>
#include <QDir>
#include <ml/ml_onnx_model.h>
#include <ml/ml_tensor.h>
#include <ml/ml_trainer.h>

Include dependency graph for inv_cmne.cpp:

Go to the source code of this file.

Detailed Description

Implementation of the CMNE solver (dSPM kernel, z-score rectification, LSTM inference, optional training driver).

SPDX-License-Identifier: BSD-3-Clause Copyright (c) 2026 MNE-CPP Authors

Author

Christoph Dinh chris.nosp@m.toph.nosp@m..dinh.nosp@m.@mne.nosp@m.-cpp..nosp@m.org

Since

2.2.0

Date

April 2026

Implements the dSPM kernel assembly, the per-source z-score rectification, the sliding-window LSTM inference via ONNX Runtime and (outside WebAssembly builds) the UTILSLIB::PythonRunner-based training driver that invokes the upstream train_cmne_lstm.py script. The training driver is a thin process wrapper — the actual PyTorch model + ONNX export lives in Python and this C++ side only forwards arguments and captures progress.

Definition in file inv_cmne.cpp.