This project is in feasibility study... I'm not even think that it will be fully working
####Possible issues
- Where do we need to place the electrode?
- Currently place around the temporal frontal lobe...
- Reference?
- Currently place around the temporal frontal lobe...
####Project details
- This project will be using the conditional GANs since we need to input
the EEG signal to generate the spectrogram of audio
- We may use the WGAN to improve the tranning stability
- The discriminator and generator model will use the same arch from EEG2Image paper.
- Then we tweaks the model later.
- [-] Finish the GAN code (follow the ThoughtViz first)
- Input : EEG Signal
- Output : An 'A' Image
- Discriminator : Discriminate between not 'A' and 'A'
- [-] BONUS : Implement CycleGAN