AudioBox Aesthetics Prediction

Upload an audio file to predict its aesthetic scores.

This demo uses the AudioBox Aesthetics model to predict aesthetic scores for audio files along 4 axes:

  • Content Enjoyment (CE)
  • Content Usefulness (CU)
  • Production Complexity (PC)
  • Production Quality (PQ)

Scores range from 0 to 10.

For more details, see the paper or code.

Aesthetic Scorest

Aesthetic Scorest
Axes name
Score
Examples