braindecode.models.SSTDPN#

class braindecode.models.SSTDPN(n_chans=None, n_times=None, n_outputs=None, input_window_seconds=None, sfreq=None, chs_info=None, n_spectral_filters_temporal=9, n_fused_filters=48, temporal_conv_kernel_size=75, mvp_kernel_sizes=None, return_features=False, proto_sep_maxnorm=1.0, proto_cpt_std=0.01, spt_attn_global_context_kernel=250, spt_attn_epsilon=1e-05, spt_attn_mode='var', activation=<class 'torch.nn.modules.activation.ELU'>)[source]#

SSTDPN from Can Han et al (2025) [Han2025].

Small Attention Convolution

SSTDPN Architecture

The Spatial-Spectral and Temporal - Dual Prototype Network (SST-DPN) is an end-to-end 1D convolutional architecture designed for motor imagery (MI) EEG decoding, aiming to address challenges related to discriminative feature extraction and small-sample sizes [Han2025].

The framework systematically addresses three key challenges: multi-channel spatial–spectral features and long-term temporal features [Han2025].

Architectural Overview

SST-DPN consists of a feature extractor (_SSTEncoder, comprising Adaptive Spatial-Spectral Fusion and Multi-scale Variance Pooling) followed by Dual Prototype Learning classification [Han2025].

  1. Adaptive Spatial-Spectral Fusion (ASSF): Uses _DepthwiseTemporalConv1d to generate a

    multi-channel spatial-spectral representation, followed by _SpatSpectralAttn (Spatial-Spectral Attention) to model relationships and highlight key spatial-spectral channels [Han2025].

  2. Multi-scale Variance Pooling (MVP): Applies _MultiScaleVarPooler with variance pooling

    at multiple temporal scales to capture long-range temporal dependencies, serving as an efficient alternative to transformers [Han2025].

  3. Dual Prototype Learning (DPL): A training strategy that employs two sets of

    prototypes—Inter-class Separation Prototypes (proto_sep) and Intra-class Compact Prototypes (proto_cpt)—to optimize the feature space, enhancing generalization ability and preventing overfitting on small datasets [Han2025]. During inference (forward pass), classification decisions are based on the distance (dot product) between the feature vector and proto_sep for each class [Han2025].

Macro Components

  • SSTDPN.encoder (Feature Extractor)

    • Operations. Combines Adaptive Spatial-Spectral Fusion and Multi-scale Variance Pooling via an internal _SSTEncoder.

    • Role. Maps the raw MI-EEG trial \(X_i \in \mathbb{R}^{C \times T}\) to the feature space \(z_i \in \mathbb{R}^d\).

  • _SSTEncoder.temporal_conv (Depthwise Temporal Convolution for Spectral Extraction)

    • Operations. Internal _DepthwiseTemporalConv1d applying separate temporal convolution filters to each channel with kernel size temporal_conv_kernel_size and depth multiplier n_spectral_filters_temporal (equivalent to \(F_1\) in the paper).

    • Role. Extracts multiple distinct spectral bands from each EEG channel independently.

  • _SSTEncoder.spt_attn (Spatial-Spectral Attention for Channel Gating)

    • Operations. Internal _SpatSpectralAttn module using Global Context Embedding via variance-based pooling, followed by adaptive channel normalization and gating.

    • Role. Reweights channels in the spatial-spectral dimension to extract efficient and discriminative features by emphasizing task-relevant regions and frequency bands.

  • _SSTEncoder.chan_conv (Pointwise Fusion across Channels)

    • Operations. A 1D pointwise convolution with n_fused_filters output channels (equivalent to \(F_2\) in the paper), followed by BatchNorm and the specified activation function (default: ELU).

    • Role. Fuses the weighted spatial-spectral features across all electrodes to produce a fused representation \(X_{fused} \in \mathbb{R}^{F_2 \times T}\).

  • _SSTEncoder.mvp (Multi-scale Variance Pooling for Temporal Extraction)

    • Operations. Internal _MultiScaleVarPooler using _VariancePool1D layers at multiple scales (mvp_kernel_sizes), followed by concatenation.

    • Role. Captures long-range temporal features at multiple time scales. The variance operation leverages the prior that variance represents EEG spectral power.

  • SSTDPN.proto_sep / SSTDPN.proto_cpt (Dual Prototypes)

    • Operations. Learnable vectors optimized during training using prototype learning losses. The proto_sep (Inter-class Separation Prototype) is constrained via L2 weight-normalization (\(\lVert s_i \rVert_2 \leq\) proto_sep_maxnorm) during inference.

    • Role. proto_sep achieves inter-class separation; proto_cpt enhances intra-class compactness.

How the information is encoded temporally, spatially, and spectrally

  • Temporal.

    The initial _DepthwiseTemporalConv1d uses a large kernel (e.g., 75). The MVP module employs pooling kernels that are much larger (e.g., 50, 100, 200 samples) to capture long-term temporal features effectively. Large kernel pooling layers are shown to be superior to transformer modules for this task in EEG decoding according to [Han2025].

  • Spatial.

    The initial convolution at the classes _DepthwiseTemporalConv1d groups parameter \(h=1\), meaning \(F_1\) temporal filters are shared across channels. The Spatial-Spectral Attention mechanism explicitly models the relationships among these channels in the spatial-spectral dimension, allowing for finer-grained spatial feature modeling compared to conventional GCNs according to the authors [Han2025]. In other words, all electrode channels share \(F_1\) temporal filters independently to produce the spatial-spectral representation.

  • Spectral.

    Spectral information is implicitly extracted via the \(F_1\) filters in _DepthwiseTemporalConv1d. Furthermore, the use of Variance Pooling (in MVP) explicitly leverages the neurophysiological prior that the variance of EEG signals represents their spectral power, which is an important feature for distinguishing different MI classes [Han2025].

Additional Mechanisms

  • Attention. A lightweight Spatial-Spectral Attention mechanism models spatial-spectral relationships

    at the channel level, distinct from applying attention to deep feature dimensions, which is common in comparison methods like ATCNet.

  • Regularization. Dual Prototype Learning acts as a regularization technique

    by optimizing the feature space to be compact within classes and separated between classes. This enhances model generalization and classification performance, particularly useful for limited data typical of MI-EEG tasks, without requiring external transfer learning data, according to [Han2025].

Parameters:

  • n_chans (int) – Number of EEG channels.

  • n_times (int) – Number of time samples of the input window.

  • n_outputs (int) – Number of outputs of the model. This is the number of classes in the case of classification.

  • input_window_seconds (float) – Length of the input window in seconds.

  • sfreq (float) – Sampling frequency of the EEG recordings.

  • chs_info (list of dict) – Information about each individual EEG channel. This should be filled with info["chs"]. Refer to mne.Info for more details.

  • n_spectral_filters_temporal (int, optional) – Number of spectral filters extracted per channel via temporal convolution. These represent the temporal spectral bands (equivalent to \(F_1\) in the paper). Default is 9.

  • n_fused_filters (int, optional) – Number of output filters after pointwise fusion convolution. These fuse the spectral filters across all channels (equivalent to \(F_2\) in the paper). Default is 48.

  • temporal_conv_kernel_size (int, optional) – Kernel size for the temporal convolution layer. Controls the receptive field for extracting spectral information. Default is 75 samples.

  • mvp_kernel_sizes (list[int], optional) – Kernel sizes for Multi-scale Variance Pooling (MVP) module. Larger kernels capture long-term temporal dependencies .

  • return_features (bool, optional) – If True, the forward pass returns (features, logits). If False, returns only logits. Default is False.

  • proto_sep_maxnorm (float, optional) – Maximum L2 norm constraint for Inter-class Separation Prototypes during forward pass. This constraint acts as an implicit force to push features away from the origin. Default is 1.0.

  • proto_cpt_std (float, optional) – Standard deviation for Intra-class Compactness Prototype initialization. Default is 0.01.

  • spt_attn_global_context_kernel (int, optional) – Kernel size for global context embedding in Spatial-Spectral Attention module. Default is 250 samples.

  • spt_attn_epsilon (float, optional) – Small epsilon value for numerical stability in Spatial-Spectral Attention. Default is 1e-5.

  • spt_attn_mode (str, optional) – Embedding computation mode for Spatial-Spectral Attention (‘var’, ‘l2’, or ‘l1’). Default is ‘var’ (variance-based mean-var operation).

  • activation (nn.Module, optional) – Activation function to apply after the pointwise fusion convolution in _SSTEncoder. Should be a PyTorch activation module class. Default is nn.ELU.

Raises:

ValueError – If some input signal-related parameters are not specified: and can not be inferred.

Notes

  • The implementation of the DPL loss functions (\(\mathcal{L}_S\), \(\mathcal{L}_C\), \(\mathcal{L}_{EF}\)) and the optimization of ICPs are typically handled outside the primary forward method, within the training strategy (see Ref. 52 in [Han2025]).

  • The default parameters are configured based on the BCI Competition IV 2a dataset.

  • The use of Prototype Learning (PL) methods is novel in the field of EEG-MI decoding.

  • Lowest FLOPs: Achieves the lowest Floating Point Operations (FLOPs) (9.65 M) among competitive SOTA methods, including braindecode models like ATCNet (29.81 M) and EEGConformer (63.86 M), demonstrating computational efficiency [Han2025].

  • Transformer Alternative: Multi-scale Variance Pooling (MVP) provides a accuracy improvement over temporal attention transformer modules in ablation studies, offering a more efficient alternative to transformer-based approaches like EEGConformer [Han2025].

Warning

Important: To utilize the full potential of SSTDPN with Dual Prototype Learning (DPL), users must implement the DPL optimization strategy outside the model’s forward method. For implementation details and training strategies, please consult the official code at [Han2025Code]: hancan16/SST-DPN

References

[Han2025] (1,2,3,4,5,6,7,8,9,10,11,12,13,14,15)

Han, C., Liu, C., Wang, J., Wang, Y., Cai, C., & Qian, D. (2025). A spatial–spectral and temporal dual prototype network for motor imagery brain–computer interface. Knowledge-Based Systems, 315, 113315.

[Han2025Code]

Han, C., Liu, C., Wang, J., Wang, Y., Cai, C., & Qian, D. (2025). A spatial–spectral and temporal dual prototype network for motor imagery brain–computer interface. Knowledge-Based Systems, 315, 113315. GitHub repository. hancan16/SST-DPN.

Methods

forward(x)[source]#

Classification is based on the dot product similarity with Inter-class Separation Prototypes (SSTDPN.proto_sep).

Parameters:

x (torch.Tensor) –

Input tensor. Supported shapes:

  • (batch, n_chans, n_times)

Return type:

Union[Tensor, Tuple[Tensor, Tensor]]

Returns:

  • logits (torch.Tensor) – If input was 3D: (batch, n_outputs)

  • Or if self.return_features is True – (features, logits) where features shape is (batch, feat_dim)

On this page