API reference#

Note

mne-denoise is under active development. Until version 1.0, the public API may evolve between releases. For reproducible analyses, record the mne-denoise version used in your work.

Names documented in this reference are public unless explicitly marked experimental. Underscore-prefixed implementation details are private.

Primary denoising API#

The main interfaces follow the artifact and method flow in the Methods guide and are listed here first.

Artifact Subspace Reconstruction#

`ASR`	Artifact Subspace Reconstruction estimator.
`AdaptiveASR`	Adaptive Artifact Subspace Reconstruction estimator.
`JugglerASR`	JugglerASR estimator with pointwise reference-sample selection.
`GuidedASR`	Guided soft-reconstruction variant of ASR.

`calibrate_asr`	Calibrate an ASR state from continuous channel-first data.
`process_asr`	Apply a calibrated ASR state to continuous data.

Sensor Noise Suppression#

SNS

Sensor Noise Suppression estimator.

compute_sns

Learn and apply Sensor Noise Suppression to channel-first data.

SOUND#

SOUND

SOUND estimator for source-informed noise suppression.

compute_sound

Compute the SOUND cleaning operator.

Spectrum interpolation#

SpectrumInterpolation

Line-noise remover based on amplitude spectrum interpolation.

interpolate_spectrum

Interpolate line-noise amplitudes in a 2-D signal spectrum.

ZapLine#

ZapLine

DSS-based line-noise removal estimator.

DSS#

Core#

`DSS`	Denoising Source Separation transformer.
`IterativeDSS`	Iterative DSS transformer.
`TimeShiftDSS`	Lag-augmented DSS estimator for repeated trials.

`compute_dss`	Compute DSS spatial filters from baseline and biased covariances.
`iterative_dss`	Extract multiple components with iterative DSS.

Biases and linear denoisers#

`LinearDenoiser`	Base class for DSS bias transformations.
`AverageBias`	Averaging bias for repeatable DSS structure.
`CycleAverageBias`	Fixed-window event-locked averaging bias.
`BandpassBias`	Bandpass-filter bias for DSS.
`LineNoiseBias`	Line-frequency bias using IIR or FFT selection.
`PeakFilterBias`	Second-order IIR peak-filter bias for DSS.
`CombFilterBias`	Comb-filter bias for harmonic frequencies.
`LagAverageBias`	Lag-averaging bias for ordinary sensor-space DSS.
`SmoothingBias`	Causal running-mean bias for DSS.
`SpectrogramBias`	Fixed STFT-mask bias for DSS.

Nonlinear denoisers#

`NonlinearDenoiser`	Base class for nonlinear DSS denoisers.
`TanhMaskDenoiser`	Scaled hyperbolic-tangent nonlinearity.
`RobustTanhDenoiser`	Residual hyperbolic-tangent nonlinearity.
`KurtosisDenoiser`	Configurable ICA contrast nonlinearity.
`SkewDenoiser`	Squared-source nonlinearity for iterative DSS.
`GaussDenoiser`	Gaussian nonlinearity for iterative DSS.
`WienerMaskDenoiser`	Local-variance Wiener mask for iterative DSS.
`VarianceMaskDenoiser`	Local-variance mask for iterative DSS.
`SpectrogramDenoiser`	STFT-mask denoiser for iterative DSS.
`DCTDenoiser`	DCT-domain denoiser.
`QuasiPeriodicDenoiser`	Cycle-template denoiser for quasi-periodic source signals.
`SmoothTanhDenoiser`	Uniformly smoothed hyperbolic-tangent nonlinearity.

Singular Spectrum Analysis#

`SingularSpectrumAnalysis`	Frequency-guided, channel-wise Basic SSA transformer.
`LocalSingularSpectrumAnalysis`	Channel-wise local-SSA transformer for high-amplitude artifact reconstruction.

`compute_basic_ssa`	Apply frequency-guided Basic SSA independently to each channel.
`compute_local_ssa`	Apply local SSA independently to each channel.

BSS-CCA#

BSSCCA

Reference-free BSS-CCA estimator.

compute_bss_cca

Learn and apply reference-free BSS-CCA to channel-first data.

iCanClean#

ICanClean

Reference-based CCA artifact-removal estimator.

compute_icanclean

Compute one iCanClean pass on continuous NumPy arrays.

SSP-SIR#

SSPSIR

Source-informed signal-space projection for TMS-evoked muscle artifact removal.

Additional API#

The secondary reference pages collect reusable building blocks and supporting interfaces: