Current ML models for music emotion recognition, while generally working quite well, do not give meaningful or intuitive explanations for their predictions. In this work, we propose a 2-step procedure to arrive at spectrogram-level explanations that connect certain aspects of the audio to interpretable mid-level perceptual features, and these to the actual emotion prediction. That makes it possible to focus on specific musical reasons for a prediction (in terms of perceptual features), and to trace these back to patterns in the audio that can be interpreted visually and acoustically.
[1]
T. Eerola,et al.
A comparison of the discrete and dimensional models of emotion in music
,
2011
.
[2]
Mohammad Soleymani,et al.
A Data-driven Approach to Mid-level Perceptual Musical Feature Modeling
,
2018,
ISMIR.
[3]
Bob L. Sturm,et al.
Local Interpretable Model-Agnostic Explanations for Music Content Analysis
,
2017,
ISMIR.
[4]
Carlos Guestrin,et al.
"Why Should I Trust You?": Explaining the Predictions of Any Classifier
,
2016,
ArXiv.
[5]
Gerhard Widmer,et al.
Towards Explainable Music Emotion Recognition: The Route via Mid-level Features
,
2019,
ISMIR.