bai-Mind-8 / README.md
eyupipler's picture
Update README.md
742fca3 verified
metadata
license: cc-by-nc-sa-4.0
language:
  - en
  - tr
pipeline_tag: text-generation
tags:
  - eeg
  - epilepsy
  - emotion
  - bci
  - seizure
  - brain
  - neuroscience
  - neuro
  - neural
  - health
  - healthcare
  - medical
  - ai
  - alzheimer
  - assistant
  - future
  - clinical
  - deep-learning
  - machine-learning
library_name: keras

bai-Mind-8 Models

Definition

bai Models are trained to recognize various tasks from brain waves. These versions are suitable for everyday use. They work with 8-channel EEG devices and are fully compatible with Neuramax.

This model family is a language model trained to transfer sentences from our minds to a computer. The model has been trained entirely for experimental purposes and is open to further development.

Performance Metrics

Model Size Parameters Accuracy Recall F1 Score Precision
bai-Mind-8 v1 1.65GB 148.23M 18% 79.3% 73.5% 68.6%

Model Applications and Informations

bai-Mind-8 v1

Definition

The bai-Mind-8 v1 model has been trained to work on 8-electrode EEG devices. It is the first of its kind in the world in the field of mind-language models. It has used the ZUCO dataset and other tokenizer models.

Warning: This model shows low accuracy. All usage details and model architecture are located in the “v1” folder.

Dataset Citation

ZuCo Dataset v1: https://osf.io/q3zws/files/

Limitations & Considerations

Important Notes:

  1. Medical Use: This models is for research purposes and should NOT replace professional medical diagnosis.
  2. Feature Dependency: Requires specific feature extraction pipeline.
  3. Patient Variability: Performance may vary across different patients.
  4. Data Privacy: EEG data must be handled with strict confidentiality.

Acknowledgments

Some of these models use Neurazum's proprietary datasets. The data is closed-source and subject to privacy regulations. Models using open-source datasets are INDICATED.

Note

Use at your own risk. Due to the complexity of EEG signals and patient variability, accuracy rates may vary. The model should be used as a supportive tool, not as a replacement for professional medical judgment. Since the data belongs to , the function structure may change in future models.

License

CC-BY-NC-SA 4.0 - see LICENSE for details.

Support


For a future that combines artificial intelligence and neuroscience! 🧠⚡

Neurazum AI Department