EEG signals used to predict music enjoyment with deep learning
EEG-Based Prediction of Music Enjoyment Using Deep Learning with Explainable Temporal Analysis
Machine LearningHuman-Computer InteractionSound
Summary
It is hard to know why people enjoy music because it changes while they listen. The authors used brainwave recordings (EEG) from 20 people along with their ratings of songs to predict how much they enjoyed the music. They tested six deep learning models and found some could estimate enjoyment better than others. They also used methods to explain which brief moments in the music influenced the predictions most. This study provides a way to predict music enjoyment across different listeners and shows how explaining AI decisions helps understand the results.
What this means in practice
- •For music technology developers: Create music recommendation systems that adapt based on users’ brain responses to songs, improving personalization beyond explicit feedback.$Commercial implications: Enables brain-powered user feedback products that customize music streaming services more effectively.
- •For brain-computer interface engineers: Incorporate enjoyment prediction into neurofeedback devices to monitor and respond to emotional engagement during music listening.
Tested on one dataset.
Authors
Parshav Singla, Simranjit Kaur, Maanya Walia, Niharika
Abstract
Understanding why listeners enjoy music is difficult because the experience is subjective and changes as a piece unfolds. Electroencephalography (EEG) offers a non-invasive way to study neural responses during listening, yet earlier work has mostly treated music emotion as a categorical label or relied on black-box models. We formulate enjoyment prediction as a regression problem on the public MUSIN-G dataset, which contains 128-channel EEG from 20 listeners together with song-level enjoyment ratings. Six deep learning models were compared under a participant-wise protocol, in which 15 listeners were used for training and validation and 5 were held out for testing. DeepConvNet, CNN-LSTM and GNN-CNN reached the lowest mean absolute error (0.97 to 0.99 on the 1 to 5 scale), while the temporal convolutional network (TCN) had the highest Pearson correlation with the ratings (r = 0.21). Gradient-based and occlusion-based explanations, averaged over all held-out windows, identify the time segments within each two-second window that most influence the TCN predictions. The study provides a reproducible cross-listener benchmark for EEG-based enjoyment regression and shows how explainability analysis complements standard metrics.