/Google_Speech_TSNE_UMAP

Primary LanguageHTML

Google_Speech_TSNE_UMAP [React.js]

Dimensionality reduction (T-SNE and UMAP) applied to selected samples from Google's Speech Commands Dataset. See it live Here

About

  • Original inspiration and method for audio processing here.
  • Code modified to render audio data with React and react-three-fiber.
  • This experiment uses the first 90 audio samples of each word from the dataset.