BrainNet

This repository contains the code and testbed for the experiments in the publication:

Fallahi, Matin, Thorsten Strufe, and Patricia Arias-Cabarcos. "BrainNet: Improving Brainwave-based Biometric Recognition with Siamese Networks." 2023 IEEE International Conference on Pervasive Computing and Communications (PerCom).

@inproceedings{fallahi2023brainnet,

  title={BrainNet: Improving Brainwave-based Biometric Recognition with Siamese Networks},

  author={Fallahi, Matin and Strufe, Thorsten and Arias-Cabarcos, Patricia},

  booktitle={2023 IEEE International Conference on Pervasive Computing and Communications (PerCom)},

  pages={53--60},

  year={2023},

  organization={IEEE}

}

Repository Structure

• Folders:

  • The "Data" folder contains preprocessed samples from the datasets described in the paper.
  • The "Plots" folder contains plots generated during the final evaluation in the "Evaluation.py".
  • The "Similarity Scores" folder contains similarity scores calculated based on different scenarios, mentioned in the paper, during training and evaluation using the "Dataset_1_2_train" or "Dataset_3_train" code.

• Files:

  • "Dataset_1_2_train.py" and "Dataset_3_train.py" contain Python files for training networks and storing similarity scores.
  • "Evaluation.py" gives you final results based on similarity scores stored previously.
  • "Network details table.pdf" represents the parameters of CNN sub-networks in BrainNet and parameters of the encompassing Siamese Network.

  Datasets Map

• P300:ERP CORE -> D_1

• N400:ERP CORE -> D_2

• P300:bi2015a -> D_3

Data

https://i62nextcloud.tm.kit.edu/index.php/s/fj7mrk8SFHi4msQ