基础模型已评测
LaBraM
5.82M 参数
A fixed configuration has been evaluated. See each task for channels and training mode.
公开 EEG 评测 · 快照 2026-09-20
每一个 EEG 分数，都附带产生它的协议。
核心矩阵覆盖 9 种解码方法、8 个固定协议、7 个公开数据集。四个部署专题另外补充了 82 项测量，涉及传感器、运动、校准与预训练。
- 协议
- 8
- 数据集
- 7
- 比较
- 39
- 方法
- 9
部署问题
解读新证据的四个角度
另外 82 项聚合测量，按它们能支持的决策来组织。它们是同一协议内的重复条件，不是 82 项独立实验，也不是总排名。
01传感器迁移
干电极与湿电极
同一批 102 名被试、两种原生八通道记录之间，解码器迁移过去会发生什么？
2 秒 SSVEP · 12 个目标 · 平衡准确率查看证据 →02运动鲁棒性移动中的解码
站立、行走与跑动时的结果，头皮与耳部记录、互不兼容的时间窗分开呈现。
SSVEP 平衡准确率 · ERP ROC AUC查看证据 →03适配预算需要多少校准？
12、24 或 48 个目标被试校准试次，对某些方法帮助更大——而且试次数不等于耗时。
共同的后续组块 · 仅用目标被试数据拟合查看证据 →04表征对照预训练有用吗？
在固定与训练集内选定两种分类头设置下，对比匹配的预训练编码器与随机初始化编码器。
两个任务 · 两种编码器 · 三次随机初始化查看证据 →全精度比例、配对对比、描述性区间、引用，以及五组各含三个随机种子的敏感性分析。
下载已审核的专题数据 · JSON ↓核心基准矩阵
9 种方法 × 8 个协议
最初的八协议快照，与上方的部署专题相互独立。空白单元格表示该方法尚未在该协议上运行——不是失败。
| 方法 | 运动想象与静息n=10 · 随机水平 50% | 空闲与指令n=4 · 检出率 | SSVEP · 8 通道n=70 · 随机水平 2.5% | SSVEP · 4 通道n=70 · 随机水平 2.5% | 心算与静息n=36 · 随机水平 50% | P300 目标 ERPn=21 · 随机水平 50% | 语义目标 ERPn=30 · 随机水平 50% | 睡眠分期n=20 · 随机水平 20% |
|---|---|---|---|---|---|---|---|---|
| CBraMod基础模型 · 8 / 8 | 62.9 | 18.3 | 33.7 | 27.6 | 62.3 | 55.0 | 55.8 | 71.3 |
| EEGNet轻量模型 · 8 / 8 | 71.0 | 1.7 | 55.8 | 44.1 | 67.6 | 53.3 | 61.4 | 56.5 |
| LaBraM基础模型 · 8 / 8 | 53.5 | 18.3 | 10.8 | 12.9 | 64.6 | 49.4≤ 随机 | 53.2 | 71.1 |
| Spectral ridge经典方法 · 7 / 8 | 54.9 | · | 50.7 | 48.2 | 56.8 | 51.0 | 51.0 | 72.2 |
| CSP+LDA经典方法 · 2 / 8 | 47.1≤ 随机 | 23.3 | · | · | · | · | · | · |
| Standard CCA经典方法 · 2 / 8 | · | · | 63.1 | 57.6 | · | · | · | · |
| Temporal ridge经典方法 · 2 / 8 | · | · | · | · | · | 53.8 | 56.3 | · |
| Deep4Net轻量模型 · 1 / 8 | · | 3.3 | · | · | · | · | · | · |
| ShallowFBCSPNet轻量模型 · 1 / 8 | · | 43.3 | · | · | · | · | · | · |
该协议最佳条长 = 在该协议随机水平与 100% 之间的位置未评测核心矩阵 · JSON ↓
条形只能纵向比较，不能横向比较：二分类任务的随机水平是 50%，睡眠分期是 20%，40 分类 SSVEP 是 2.5%，而且每个协议的队列和电极布局都不同。不存在总分。空闲一列只报告指令检出率——请结合该协议中的误触发率一起读。
逐个协议
每个分数背后的证据
队列、电极布局、训练预算、来源条款与已知局限，和它们所属的数字放在一起。
协议步骤、局限说明、模型备注与数据集署名来自发布数据本身，保持英文原文——它们随数据一起被审核，翻译会让网页与可下载文件不再一致。
Rest versus right-hand imagery
ds003810 · 10 名被试 · 1,200 epochs · 5 participant-disjoint folds
| 模型 / 配置 | 平衡准确率 | 宏平均 F1 | 计算耗时 |
|---|---|---|---|
| Supervised fit · 15 ch | 54.9%Descriptive 95% interval: 53.2–56.7% | 0.544Mean across held-out participants | 0.2s |
| Supervised fit · 15 ch | 47.1%Descriptive 95% interval: 43.8–50.2%不高于 50% 随机水平 | 0.406Mean across held-out participants | 8.8s |
| Frozen encoder + ridge head · 15 ch | 53.5%Descriptive 95% interval: 51.4–55.4% | 0.509Mean across held-out participants | 2.2s |
| Frozen encoder + ridge head · 15 ch | 62.9%Descriptive 95% interval: 59.6–66.5% | 0.617Mean across held-out participants | 5.0s |
| Scratch · 20 epochs · 15 ch | 71.0%Descriptive 95% interval: 65.0–76.3%仅单个随机种子——为 3 次运行中最高的一次（68.17–71.00%，均值 69.47%） | 0.680Mean across held-out participants | 55.5s |
横向滚动表格查看其余列。
固定预算与适配器；不存在通用排名。评分耗时包括拟合与预测，可能含加速器等待时间，不含数据准备。
谨慎解读
Ten-person laboratory task with prompted rest. This is not continuous-idle monitoring or a physical low-channel headset test.
方法 →模型目录
从轻量基线到基础模型
可获得性与实测表现是两回事。参数量取决于主干网络与任务配置。
基础模型权利审查中
EEGPT
25.29M 参数
Evaluated locally, but no score is published in this release: the checkpoint license is unresolved. Results are withheld pending that review, not because the model failed to run.
基础模型已评测
CBraMod
4.88M 参数
A fixed configuration has been evaluated. See each task for channels and training mode.
轻量模型已评测
EEGNet
1.9K 参数
A fixed configuration has been evaluated. See each task for channels and training mode.
轻量模型已评测
ShallowFBCSPNet
30.5K 参数
A fixed configuration has been evaluated. See each task for channels and training mode.
轻量模型已评测
Deep4Net
274.6K 参数
A fixed configuration has been evaluated. See each task for channels and training mode.
经典方法已评测
CSP+LDA
— 参数
A fixed configuration has been evaluated. See each task for channels and training mode.
基础模型需要适配器
BIOT
3.19M 参数
Pretrained bipolar tokens do not directly match the unipolar recordings. A separately validated adapter is required.
基础模型访问受限
REVE Base
— 参数
Code and electrode positions are accessible. Base weights require accepting the official access agreement; they have not been downloaded.
基础模型研究候选
Signal-JEPA
3.46M 参数
Public, ungated weights; research candidate only; not downloaded or executed for this release. The official model card provides a 13,847,488-byte safetensors checkpoint trained on Lee2019 at 128 Hz. All 17 ds005342 channel names match its 62-channel pretraining layout, but a separate locked adaptation protocol is still required.
基础模型研究候选
BENDR
157.00M 参数
Public, ungated weights; research candidate only; not downloaded or executed in this expansion. The Braindecode checkpoint is 628,580,476 bytes and expects 20 channels at 250 Hz. A documented montage or input adapter is required for 17-channel data.
基础模型访问未核实
CSBrain
— 参数
Public source and an official Google Drive weight link are listed; anonymous weight download, file size, and execution remain unverified. The backbone accepts caller-supplied brain regions and channel ordering, while the published BCIC head is fixed to 22 channels. A new adapter and a checkpoint-key loading audit are required.
基础模型研究候选
Neuro-GPT
— 参数
Public code and Hugging Face weights; research candidate only; not downloaded or executed for this release. The model page shows an approximately 318 MB checkpoint. The published EEGConformer input path is fixed to 22 channels, so using 17 channels requires an explicit mapping or replacement of the spatial layer.
轻量模型研究候选
EEG-Conformer
— 参数
Public source; no official pretrained checkpoint was found; not executed for this release. This is a from-scratch baseline rather than a foundation-model checkpoint. The author script fixes the spatial convolution to 22 channels and must be parameterized for other montages.
轻量模型研究候选
EEGSimpleConv
— 参数
Public implementation; no pretrained weights; not executed for this release. The maintained Braindecode implementation accepts the channel count and sampling rate directly, making it a practical lightweight from-scratch comparator.
经典方法已评测
Standard CCA
— 参数
Fixed reference method; inspect each task for input features and fit protocol.
经典方法已评测
Spectral ridge
— 参数
Fixed reference method; inspect each task for input features and fit protocol.
经典方法已评测
Temporal ridge
— 参数
Fixed reference method; inspect each task for input features and fit protocol.
公开数据
公开数据，有据可查的实验
来源条款与数值结果分开审查。尚未厘清的数据不纳入本次发布。
运动想象 / 静息
ds003810 ↗
Clear CC0 snapshot plus dataset-specific ethics and signed-consent evidence; aggregate-only publication materially limits privacy exposure.
- 被试
- 10
- 通道
- 15
已发布聚合结果CC0-1.0 ↗Peterson et al. · OpenNeuro ds003810, version 2.0.2. Study: https://pmc.ncbi.nlm.nih.gov/articles/PMC9114495/
心算 / 静息
EEGMAT ↗
Attribution license and study-specific approval/consent are documented; eligibility assumes aggregate-only output and exclusion of subject-info fields.
- 被试
- 36
- 通道
- 19
已发布聚合结果Open Data Commons Attribution License 1.0 ↗Igor Zyma, Ivan Seleznov, Anton Popov, Mariia Chernykh, Oleksii Shpenkov · EEG During Mental Arithmetic Tasks 1.0.0, PhysioNet, doi:10.13026/C2JQ1P. Study: Zyma et al. (2019), doi:10.3390/data4010014. PhysioNet platform: Pollard et al. (2026), doi:10.1038/s44360-026-00096-z.
五期睡眠分期
EESM19 scalp subset ↗
Clear upstream CC0 and study ethics evidence; use only the pinned scalp-signal derivative and publish aggregates, with provenance disclosed.
- 被试
- 20
- 通道
- 6
已发布聚合结果CC0-1.0 declared by upstream and mirror ↗Kaare B. Mikkelsen et al. · Accurate whole-night sleep monitoring with dry-contact ear-EEG (2019), doi:10.1038/s41598-019-53115-3; OpenNeuro ds005185 v1.0.2. Processed mirror: Zachary1150/EESM19-Processed.
40 目标 SSVEP
BETA ↗
Eligible only under the stated personal, research-led, noncommercial operation. Any ads, sponsorship, fees, or use directed toward commercial advantage requires fresh review or permission.
- 被试
- 70
- 通道
- 4 / 8 selected
已发布聚合结果CC BY 4.0 ↗Bingchuan Liu et al. · BETA: A Large Benchmark Database Toward SSVEP-BCI Application (2020), doi:10.3389/fnins.2020.00627. Figshare 12264401 v3; mirror Bingchuan/BETA.
研究来源
BNCI2014-008 ↗
Small clinical cohort and linked health attributes create disproportionate reidentification risk; aggregate-only output does not resolve the missing secondary-publication consent evidence.
- 被试
- —
- 通道
- —
未纳入本次发布CC BY-NC-ND 4.0 ↗BNCI Horizon 2020 catalog · Original researchers credited at source.
研究来源
BNCI2014-009 ↗
License permits only noncommercial unadapted redistribution, and the consent/ethics chain for public secondary results remains unverified.
- 被试
- —
- 通道
- —
未纳入本次发布CC BY-NC-ND 4.0 ↗BNCI Horizon 2020 catalog · Original researchers credited at source.
研究来源
stew-monster-processed ↗
Ethics evidence is positive, but upstream license and the processed mirror's authority to apply CC BY remain unresolved.
- 被试
- —
- 通道
- —
未纳入本次发布Mirror declares CC BY 4.0; exact upstream DataPort license not independently verified ↗monster-monash/STEW Hugging Face processed mirror · Original researchers credited at source.
语义目标 ERP
TMNRED / ds005383 ↗
Strong study-specific open-sharing evidence. Aggregate metrics avoid redistribution of stimulus text and participant metadata; credit under CC BY.
- 被试
- 30
- 通道
- 30
已发布聚合结果OpenNeuro metadata says CC0; accompanying publication/GitHub says CC BY 4.0 ↗Yanru Bai, Qi Tang et al. · TMNRED, A Chinese Language EEG Dataset for Fuzzy Semantic Target Identification in Natural Reading Environments (2025), doi:10.1038/s41597-025-05036-2. OpenNeuro ds005383 v1.0.0.
P300 目标 ERP
ds006593 ↗
Pinned CC0 release with dataset-specific IRB and consent evidence and comparatively low metadata exposure.
- 被试
- 21
- 通道
- 19
已发布聚合结果CC0-1.0 ↗OpenNeuro ds006593 contributors · version 1.0.0, doi:10.18112/openneuro.ds006593.v1.0.0; original author credits retained at the linked source.
研究来源
physionet-eegmmidb-1.0.0 ↗
Reuse license is clear, but the dataset-specific human-subject consent and ethics chain is not evidenced strongly enough for a new public benchmark release.
- 被试
- —
- 通道
- —
未纳入本次发布Open Data Commons Attribution License 1.0 ↗PhysioNet EEG Motor Movement/Imagery Dataset 1.0.0 · Original researchers credited at source.
研究来源
BNCI2014-001 ↗
Per-dataset license is verified, but consent evidence and the application of ND to the planned publication pipeline should be clarified.
- 被试
- —
- 通道
- —
未纳入本次发布CC BY-ND 4.0 ↗BNCI Horizon 2020 catalog · Original researchers credited at source.
研究来源
BNCI2014-004 ↗
Per-dataset license is verified, but consent evidence and the application of ND to the planned publication pipeline should be clarified.
- 被试
- —
- 通道
- —
未纳入本次发布CC BY-ND 4.0 ↗BNCI Horizon 2020 catalog · Original researchers credited at source.
提示同步的空闲 / 指令
ds005342 ↗
Eligible only as an aggregate-only case study: remove all participant IDs/results and condition breakdowns, label n=4 prominently, avoid subgroup claims, and make no claim that the cohort is anonymous.
- 被试
- 4 evaluated / 32 prepared
- 通道
- 17
已发布聚合结果CC0-1.0 ↗OpenNeuro ds005342 contributors · version 1.0.3, doi:10.18112/openneuro.ds005342.v1.0.3; associated study doi:10.3389/fninf.2022.961089. Original author credits are retained at the linked source.
领域动态
研究前沿动态
精选研究动态，均链接至原始来源。
Benchmark
A new benchmark compares EEG foundation models with task-specific decoders ↗
Researchers surveyed 55 representative models and evaluated 12 open-source foundation models with task-specific baselines across 13 datasets and nine BCI paradigms. The study reports that linear probing is often insufficient and that larger models do not automatically generalize better, reinforcing the need for fixed cross-subject and few-shot calibration protocols.
Original arXiv paperModel release
REVE releases an EEG foundation model for varying electrode layouts ↗
REVE reports pretraining on 92 datasets, about 60,000 hours of EEG, and 25,000 participants, with position encodings designed for different electrode arrangements. Its public code, weights, and tutorials make cross-montage transfer an auditable evaluation target. Access to the Base checkpoint is gated; it is not yet available in our local test pool.
Original arXiv paper and NeurIPS 2025 paper pageResearch release
NeurIPT brings heterogeneous EEG configurations into a unified pretraining framework ↗
NeurIPT models homogeneous and heterogeneous spatiotemporal relationships across BCI tasks and electrode layouts and appeared as a NeurIPS 2025 poster. Its official repository still says that code restoration is in progress, so it is a research update rather than a reproducible model card.
OpenReview / NeurIPS 2025Model release
CSBrain adds cross-scale spatiotemporal structure to general EEG decoding ↗
CSBrain combines local temporal windows, brain regions, and structured sparse attention, with reported experiments spanning 16 datasets and 11 task types. The official repository provides pretraining and downstream scripts plus a public weight entry point, making it a candidate for a later unified-protocol evaluation.
Original arXiv paper and NeurIPS 2025 paper pageReal-time BCI
An eight-channel, low-preprocessing self-supervised EEG model targets real-time BCI ↗
This study applies HuBERT-style self-supervised learning to eight-channel scalp EEG and examines motor imagery, P300, subject differences, and alpha rhythms. It motivates a separate deployment-oriented track for low channel counts, limited preprocessing, and runtime budgets.
Original arXiv paper先有证据，再谈排名
分数只有连同条件才有用。
Only explicitly reviewed cohort-level research results are exported. Raw EEG, individual results and model weights stay outside the website. Each comparison states its source, license, protocol and limitations. No cross-task overall score.
下载全部结果 · JSON ↓能跑通前向传播不等于基准测试
下载、加载检查与完成评分的评测有各自的状态标签。仍待适配器的模型没有结果。
把失败模式摆出来
报告漏检、误触发与拒识。短时间内零错误不能证明全天可靠。
比较同一个任务
数据划分、通道、预处理与训练模式都随协议一起给出。冻结编码器与从头训练的基线分开标注。
从固定、经审计的本地运行做起
实验在 Mac 或本地 GPU 工作站上离线运行。耗时只对该配置有效。本网站不做任何 EEG 推理或诊断。