Music auto-tagging is key for improving music discovery and recommendation. Existing models in Music Information Retrieval (MIR) struggle with real-world noise. This study proposes a method that integrates Domain Adversarial Training (DAT) into the music domain, enabling robust music representations that withstand noise. Adding various synthesized noisy music data improves the model’s generalization across different noise levels. The proposed architecture demonstrates enhanced performance in music auto-tagging by effectively utilizing unlabeled noisy music data.
Publication date: 31 Jan 2024
Project Page: Not Provided
Paper: https://arxiv.org/pdf/2401.15323
