Published on

MediaPipe with a bit of Bard


I was asked to give one of the Machine Learning talks at the Singapore Google I/O Extended 2023 event held in the Singapore Google Developer Space.

My talk was titled "MediaPipe with a bit of Bard", and included 4 different demos of MediaPipe functionality:

  • Audio classification (showing the Google instant demo page)
  • Facial landmarks (using the Javascript version via Codepen)
  • Interactive Segmentation (using the Python notebook, migrated to Colab, and using Bard to fix the foreground/background mistake in Google's code)
  • Text Classification (showing how TFLite could be used to ~minify MobileBERT, and Bard could be used to generate sample data)

And the slides for my talk are here :

Presentation Screenshot

If there are any questions about the presentation please ask below, or contact me using the details given on the slides themselves.

Presentation Screenshot