/244-project

Andy, Ming, and Zhengkai's CSE 244 final project repo

Primary LanguagePython

setup

  • install tensorflow (we used version 2.1.0)
  • pip install -r requirements.txt

get the data

download the TIMIT dataset from [this kaggle url] (kaggle.com/mfekadu/darpa-timit-acousticphonetic-continuous-speech) and unzip it into a directory named "data/timit" inside the root of this repository.

run

python noisy2txt.py

configuring the run

You can adjust options for the run by editing (or adding entries to) the custom_static_params dictionary at the bottom of "noisy2txt.py".

You can also run stft_denoise.py to train a larger denoisng model. A larger dataset need to be downloaded.

The model will be saved under the model directory and you can evaluate it with sftf_eval.py.

To listen to dataset as it will be fed to the model

python timit_loader.py Note that this allows you to check that the transcripts are cropped properly and that the noise level is as desired.