Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Verified mel spectrograms in winml sample against pytorch #400

Open
wants to merge 10 commits into
base: master
Choose a base branch
from

Conversation

janezdu
Copy link
Contributor

@janezdu janezdu commented Jul 23, 2021

Main changes included in this PR involve:

  • Tuned the mel spectrogram settings to match more closely with PyTorch. Rendering the tensor data from a learning model in python/librosa shows that the learning model creates a mel spectrogram very similar to the ones made by torchaudio.
  • UI controls for the settings of the melspectrogram added
  • Resampling via learning model added
  • Some refactoring in PreprocessModel.cs; methods are now "less static" and include information about the audio file/mel spectrogram settings stored as instance properties.
  • Flip and transpose mel spectrogram in learning model so that time axis goes left to right and Mel matrix goes low to high (up to down)

@janezdu janezdu requested a review from a team as a code owner July 23, 2021 17:53
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

1 participant