Feature centric based deep learning approach for music mood recognition with HuBERT transformer model
Yaoyu Sun · 2025
Read Feature centric based deep learning approach for music mood recognition with HuBERT transformer model by Yaoyu Sun (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/pmc-12749982.txt.
Description
Yaoyu Sun, writing alone in twenty twenty-five, built a deep learning system that recognizes the mood of music. Knowing a song's emotional tone helps recommendation engines, mental health monitoring, and emotionally aware software. Sun compared a transformer model called HuBERT against other approaches, including ConvFormer, LSTM networks, and the pre-trained YAMNet. The dataset held five mood classes, from aggressive to romantic, with five hundred audio files in each.
The feature-centric method draws on several audio representations, including spectrogram-based and cepstral features. Combining rich features with the transformer's sequence modeling is the paper's core idea. Sun evaluates which model reads musical emotion most accurately. Work like this underpins systems that sort huge music libraries by feeling rather than by genre or artist alone.
Catalog record
| CREATOR | Yaoyu Sun |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 111 KB |
| VERSION | 1.0 |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |