Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2509.04980 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866916935637139456 |
|---|---|
| author | Liu, Yuxuan Zhang, Peihong Sang, Rui Li, Zhixin Li, Shengchen |
| author_facet | Liu, Yuxuan Zhang, Peihong Sang, Rui Li, Zhixin Li, Shengchen |
| contents | Music adversarial attacks have garnered significant interest in the field of Music Information Retrieval (MIR). In this paper, we present Music Adversarial Inpainting Attack (MAIA), a novel adversarial attack framework that supports both white-box and black-box attack scenarios. MAIA begins with an importance analysis to identify critical audio segments, which are then targeted for modification. Utilizing generative inpainting models, these segments are reconstructed with guidance from the output of the attacked model, ensuring subtle and effective adversarial perturbations. We evaluate MAIA on multiple MIR tasks, demonstrating high attack success rates in both white-box and black-box settings while maintaining minimal perceptual distortion. Additionally, subjective listening tests confirm the high audio fidelity of the adversarial samples. Our findings highlight vulnerabilities in current MIR systems and emphasize the need for more robust and secure models. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2509_04980 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | MAIA: An Inpainting-Based Approach for Music Adversarial Attacks Liu, Yuxuan Zhang, Peihong Sang, Rui Li, Zhixin Li, Shengchen Sound Machine Learning Audio and Speech Processing Music adversarial attacks have garnered significant interest in the field of Music Information Retrieval (MIR). In this paper, we present Music Adversarial Inpainting Attack (MAIA), a novel adversarial attack framework that supports both white-box and black-box attack scenarios. MAIA begins with an importance analysis to identify critical audio segments, which are then targeted for modification. Utilizing generative inpainting models, these segments are reconstructed with guidance from the output of the attacked model, ensuring subtle and effective adversarial perturbations. We evaluate MAIA on multiple MIR tasks, demonstrating high attack success rates in both white-box and black-box settings while maintaining minimal perceptual distortion. Additionally, subjective listening tests confirm the high audio fidelity of the adversarial samples. Our findings highlight vulnerabilities in current MIR systems and emphasize the need for more robust and secure models. |
| title | MAIA: An Inpainting-Based Approach for Music Adversarial Attacks |
| topic | Sound Machine Learning Audio and Speech Processing |
| url | https://arxiv.org/abs/2509.04980 |