SeongYeonPark

For this stretch factor, the stretch effect has better performance.

1

When running the below code, ``` from scipy.io.wavfile import read import sox import numpy as np path = 'input.wav' sr, wav = read(path) tfm = sox.Transformer() tfm.set_globals(verbosity=0) stretch_ratio = np.random.normal(1.05,0.125)...

I have a question about the perplexity term (in the VQ-VAE).

2

As far as I understood, the perplexity used in this repo's VQ-VAE is kind of "meaningfully used codebook token numbers". When only one codebook token is used, perplexity is 1....

I have a question about your WER/CER results in the paper.

2

In your paper, you report WER and CER results of about 4.23% and 1.46%. Also, you mentioned that you used https://huggingface.co/facebook/hubert-large-ls960-ft as the ASR model. But, when using the same...

Question about the scale of gaussian noise in the mel spectrogram

3

In the proposed SR-augmentation, when the mel is squeezed, you pad it with the highest frequency bin value and add Gaussian noise. Can you share the scale of the added...