DavidVivancos/MindBigData2022
MindBigData 2022 is a large-scale EEG signal dataset comprising three primary datasets and their sub-datasets. The data were collected using various EEG devices such as MindWave, EPOC1, Muse1, Insight1, etc., with detailed sampling rates and channel information. The dataset is split into 80% training and 20% testing, containing both labels and EEG recordings. Each sub-dataset has specific device and sampling‑rate configurations. Specifically: 1. MindBigData MNIST of Brain Digits – four sub-datasets based on MindWave, EPOC1, Muse1, and Insight1; 2. MindBigData Imagenet of the Brain – two sub-datasets based on Insight1 EEG signals and spectrograms; 3. MindBigData Visual MNIST of Brain Digits – three sub-datasets based on Muse2, Cap64, and Cap64 Morlet devices.
Dataset description and usage context
MindBigData 2022 A Large Dataset of Brain Signals
1. MindBigData MNIST of Brain Digits
- Data source: http://mindbigdata.com/opendb/index.html
- Data split: 80% Train, 20% Test
- Data processing: EEG signals are resampled to match the original headset sampling rate, include head information, and the simplified data contain only labels and EEG recordings
- Sub‑datasets:
- MindWave1: 1 EEG Channel, 1024 samples × Channel
- EPOC1: 14 EEG Channels, 256 samples × Channel
- Muse1: 4 EEG Channels, 440 samples × Channel
- Insight1: 5 EEG Channels, 256 samples × Channel
2. MindBigData Imagenet of the Brain
- Data source: http://mindbigdata.com/opendb/imagenet.html
- Data split: 80% Train, 20% Test
- Data processing: Includes ILSVRC2013 class labels, one‑hot name list, RGB pixel values resampled to 150×150, and EEG data
- Sub‑datasets:
- Insight1 EEG: 5 EEG Channels, 384 samples per channel
- Insight1 Spectrogram: 64×64‑pixel spectrograms replace EEG
3. MindBigData Visual MNIST of Brain Digits
- Data source: http://mindbigdata.com/opendb/visualmnist.html
- Data split: 80% Train, 20% Test
- Data processing: Includes labels, original MNIST pixels of 28×28, and EEG data
- Sub‑datasets:
- Muse2: 5 EEG Channels, 3 PPG Channels, 3 ACC Channels, 3 GYR Channels, 512 samples × Channel
- Cap64: 64 EEG Channels, 400 samples × Channel
- Cap64 Morlet: 64 EEG Channels, 400 samples × Channel, uses Morlet PNG images as EEG output
Pair the dataset with AI analysis and content workflows.
Once the source passes your review, move straight into summarization, transformation, report drafting, or presentation generation with the JuheAI toolchain.