Skip to content

Commit

Permalink
adapted to python only phase recovery implementation
Browse files Browse the repository at this point in the history
  • Loading branch information
andimarafioti authored Jul 31, 2019
1 parent 5bf156c commit 7f039f0
Showing 1 changed file with 1 addition and 6 deletions.
7 changes: 1 addition & 6 deletions README.md
Original file line number Diff line number Diff line change
@@ -76,12 +76,7 @@ To generate the magnitudes from TiFGAN-M , please use:
cd specgan/generate_commands
python 64md_8k.py
```
Then, the signals can be reconstructed in MATLAB/octave with the following scripts `recover_phase_from_mags.m` or `recover_phase_from_mags_and_derivs.m`. Alternatively, for MATLAB you can try the following one-liner command:
```
matlab -nodesktop -nosplash -nodisplay -r \
"try, run('recover_phase_from_mags.m'), catch, exit(1), end, exit(0);"
```
This command will work only if the function ltfatstart is in the path of MATLAB/octave.
The output of that script is a .mat file containing the generated spectrograms and the reconstructed sounds.

## Pre-trained networks
The checkpoints used for the evaluation of the [paper][paper] can be downloaded [here](https://zenodo.org/record/2562819). Please extract the archiv in the folder `saved_results`. To generate magnitudes using those checkpoints, use on of the following commands:

0 comments on commit 7f039f0

Please sign in to comment.