Re:Reproducible machine learning with PyTorch and Quilt
by stealthcat on Tuesday July 17, 2018 @06:25PM
(#17554018)
Why don't you use STFT + Conv2D like Deep Speech 2 did.
It works well in my case.
Re:Reproducible machine learning with PyTorch and Quilt
by p1esk on Wednesday July 18, 2018 @12:16AM
(#17555545)
The DeepSpeech2 paper does not include any details about audio processing. I see an older Baidu-Research implementation of DS1 that uses "log of linear spectrogram from FFT energy". Also, there's a pytorch implementation [1], where they use Librosa's STFT, is that what you're referring to?
That's two more implementations that I haven't considered. I'm sure most of the processing steps under the hood are the same or similar, but as I'm not an audio processing expert, I can't tell which method is better (and why).
And it's hard to tell if it "works well" because or despite the way I processed the files.
Not to mention every one of those implementation packages their preprocessed version into a different data format, and then creates a different data pipeline (and I only looked at tensorflow implementations)