BRAIN2SPEECH

This repository contains the BRAIN2SPEECH Home assigment that was created for the BME Deep Learning course.

The members of the team

System requiments for running the scripts and notebook:

data_prepare.py : This file contains helper methods for data preparation.
HF_MS1.ipynb : This Jupyter notebook downloads the data from the server and create train, validation and test datasets from it. This needs to be run, in order to train the neural nets using script HF_seperate_models.ipynb
HF_MS2.ipynb: This Jupyter notebook creates a simple CNN network, trainning this network and evaluate it using MAE metrics
HF_final.ipynb :This Jupyter notebook contains the final models. Running it is straightforward as it was created specifically to create the neural nets with the best prediction performance. The blocks need to be run in order, and the end results are going to be put into the folders 'audio'.
./scripts : This directory contains scripts that was created by NeuralinterfacingLab. We used it for data extraction.
./audio : This directory contains the synthesized audio files.
./logs :This directory contains the logs of tested models.
Speech_Synthesis.pdf :The documentation

Name		Name	Last commit message	Last commit date
Latest commit History 65 Commits
audio		audio
logs		logs
scripts		scripts
.gitignore		.gitignore
Beszéd_szintézis_absztrakt_hun.pdf		Beszéd_szintézis_absztrakt_hun.pdf
Beszédszintézis EEG agyi felvételekből.pptx		Beszédszintézis EEG agyi felvételekből.pptx
Beszédszintézis EEG agyi felvételekből_eloadas.pptx		Beszédszintézis EEG agyi felvételekből_eloadas.pptx
HF_MS1.ipynb		HF_MS1.ipynb
HF_MS2.ipynb		HF_MS2.ipynb
HF_final.ipynb		HF_final.ipynb
HF_fn-dnn.ipynb		HF_fn-dnn.ipynb
HF_merged.ipynb		HF_merged.ipynb
HF_separate_models.ipynb		HF_separate_models.ipynb
README.md		README.md
Speech_Synthesis.pdf		Speech_Synthesis.pdf
Speech_synthesis_from_sEEG_via_DL.pdf		Speech_synthesis_from_sEEG_via_DL.pdf
Speech_synthesis_latex.zip		Speech_synthesis_latex.zip
data_prepare.py		data_prepare.py