braindecode/signal-jepa_without-chans

Signal-JEPA

Self-supervised pre-trained weights for the Signal-JEPA foundation model from Guetschel et al. (2024), packaged for use with braindecode. See the full API reference in the docs: braindecode.models.SignalJEPA.

The model was pre-trained on the Lee2019 dataset (62 EEG channels in the 10-10 layout, sampled at 128 Hz). The repo ships the weights together with a config.json so they can be loaded in one line with YourModelClass.from_pretrained(repo_id, ...).

Available checkpoints

Two variants are published:

repo ID	channel embedding included	when to use
braindecode/signal-jepa	✓ 62-row _ChannelEmbedding aligned with the pre-training layout	your recording channels are a subset (by name, case-insensitive) of the 62 pre-training channels — you want to reuse the learned spatial embeddings
braindecode/signal-jepa_without-chans	✗ only the SSL backbone (feature encoder + transformer)	your channels are not a subset of the pre-training set, or you prefer to train channel embeddings from scratch

If you are unsure, start with braindecode/signal-jepa_without-chans: it always works, regardless of your electrode layout.

Training data

The data that shaped these weights. The ids are the datasets: field of this card, which braindecode reads into model.card_data.datasets when you call from_pretrained (MOABB class names for MOABB datasets, OpenNeuro / NEMAR accessions, Hub ids otherwise). nm/on ids are datasets on NEMAR (on = NEMAR's copy of the OpenNeuro dataset with the same number). Rows marked private: are data the authors did not release; they are not in datasets:.

datasets: id	Dataset	Stage	Modality
nm000338	Lee et al. 2019 OpenBMI, motor imagery paradigm (54 subjects, 62 EEG channels), MOABB Lee2019_MI	pretraining	EEG
nm000323	Lee et al. 2019 OpenBMI, ERP (P300 speller) paradigm, MOABB Lee2019_ERP	pretraining	EEG
nm000273	Lee et al. 2019 OpenBMI, SSVEP paradigm, MOABB Lee2019_SSVEP	pretraining	EEG

• S-JEPA self-supervised pre-training (16 s examples, 60 % spatial mask, run adeuwv4s) on Lee2019 subjects 1-40, all three paradigms pooled, no labels; subjects 41-47 = SSL validation set used for early stopping / best-checkpoint selection; subjects 48-54 not used for these weights (downstream evaluation only). All 179 tensors byte-identical to the original file.
• Sources: arXiv:2403.11772 (HTML), section DATASETS; https://huggingface.co/api/models/braindecode/signal-jepa/commits/main (commit 9e20126cf5); braindecode check: tensor-byte sha256 check -> compare_sjepa.json; W/bd/braindecode/models/signal_jepa.py.

Quick start

Base model (pre-training architecture)

The base model outputs contextual features, not class predictions. Use it for downstream feature extraction or further SSL.

from braindecode.models import SignalJEPA

# With the pre-trained channel embeddings (recording channels ⊂ pre-train set):
model = SignalJEPA.from_pretrained("braindecode/signal-jepa")

# Or: with your own channels, kept aligned to the pre-training embedding table
model = SignalJEPA.from_pretrained(
    "braindecode/signal-jepa",
    chs_info=raw.info["chs"],           # subset of the 62 pre-training channels
    channel_embedding="pretrain_aligned",
)

# Or: without pre-trained channel embeddings (any electrode layout):
model = SignalJEPA.from_pretrained(
    "braindecode/signal-jepa_without-chans",
    chs_info=raw.info["chs"],
    strict=False,  # the channel-embedding weight is intentionally missing
)

Downstream architectures

Three classification architectures are introduced in the paper:

• a) Contextual — uses the full transformer encoder
• b) Post-local — discards the transformer; spatial convolution after local features
• c) Pre-local — discards the transformer; spatial convolution before local features

All three add a freshly-initialized classification head on top of the SSL backbone. The head is not part of the checkpoint and will be trained from scratch during fine-tuning; pass strict=False so from_pretrained does not complain about those missing keys.

from braindecode.models import (
    SignalJEPA_Contextual,
    SignalJEPA_PreLocal,
    SignalJEPA_PostLocal,
)

# a) Contextual — keeps the transformer
model = SignalJEPA_Contextual.from_pretrained(
    "braindecode/signal-jepa",          # or "signal-jepa_without-chans"
    n_times=256,                         # e.g. 2 s at 128 Hz
    n_outputs=4,
    strict=False,                        # ignore un-trained classification head
)

# b) Post-local — transformer discarded
model = SignalJEPA_PostLocal.from_pretrained(
    "braindecode/signal-jepa_without-chans",
    n_chans=19,
    n_times=256,
    n_outputs=4,
    strict=False,
)

# c) Pre-local — transformer discarded
model = SignalJEPA_PreLocal.from_pretrained(
    "braindecode/signal-jepa_without-chans",
    n_chans=19,
    n_times=256,
    n_outputs=4,
    strict=False,
)

See the braindecode tutorial Fine-tuning a Foundation Model (Signal-JEPA) for a complete example including layer freezing and training with skorch.EEGClassifier.

Channel embedding modes

SignalJEPA and SignalJEPA_Contextual accept a channel_embedding kwarg:

• "scratch" (default): the _ChannelEmbedding table has one row per user channel, initialized from chs_info. Compatible with the without-chans checkpoint.
• "pretrain_aligned": the table has 62 rows in the pre-training order, forward indexes into the subset matching your chs_info (matched by channel name, case-insensitive). Compatible with the full checkpoint.

from_pretrained picks the right mode automatically based on the checkpoint's config.json; override with the channel_embedding= kwarg if needed.

Citation

@article{guetschel2024sjepa,
  title   = {S-JEPA: towards seamless cross-dataset transfer
             through dynamic spatial attention},
  author  = {Guetschel, Pierre and Moreau, Thomas and Tangermann, Michael},
  journal = {arXiv preprint arXiv:2403.11772},
  year    = {2024},
}

Cite Braindecode

If you use this braindecode port in scientific work, cite the software on Zenodo, doi:10.5281/zenodo.17699192, and the original model and datasets cited above.

@software{braindecode,
  author = {Aristimunha, Bruno and
            Guetschel, Pierre and
            Wimpff, Martin and
            Gemein, Lukas and
            Rommel, Cedric and
            Banville, Hubert and
            Sliwowski, Maciej and
            Wilson, Daniel and
            Brandt, Simon and
            Gnassounou, Théo and
            Paillard, Joseph and
            {Junqueira Lopes}, Bruna and
            Sedlar, Sara and
            Moreau, Thomas and
            Chevallier, Sylvain and
            Gramfort, Alexandre and
            Schirrmeister, Robin Tibor},
  title = {Braindecode: toolbox for decoding raw electrophysiological brain data
           with deep learning models},
  url = {https://github.com/braindecode/braindecode},
  doi = {10.5281/zenodo.17699192},
  publisher = {Zenodo},
  license = {BSD-3-Clause},
}