audioeeg / README.md
inogii's picture
Create README.md
8e65f89 verified
metadata
license: mit
language:
  - en

AudioEEG

Dataset for the eeg2speech project training. It is divided into 3 different data types:

  • Audio Wav Recordings: corresponding to chunks of the brennan2019 dataset wav recordings. They are chunks of an audiobook of Alice in Wonderland.
  • Audio Embeddings: embeddings of the audio chunks using Wav2Vec2.0
  • EEG Embeddings: embeddings of the eeg signals of subject0 in brennan2019 dataset, for the corresponding audio chunk. Obtained with the encoder in brainmagick