# Distinguishing AI-Generated Music from Edited Audio as a Hard-Negative Robustness Task

arXiv: https://arxiv.org/abs/2608.14916 · HTML: https://mlsanity.com/paper/2608.14916/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 13 of its 18 references that are in the map.
- First posted on arXiv 14 Aug 2026.

## Builds on (most cited first, 13 of 13)

- [AST: Audio Spectrogram Transformer](https://mlsanity.com/paper/2104.01778/) (2021)
- [PANNs: Large-Scale Pretrained Audio Neural Networks for Audio Pattern Recognition](https://mlsanity.com/paper/1912.10211/) (2019)
- [Simple and Controllable Music Generation](https://mlsanity.com/paper/2306.05284/) (2023)
- [The Creation and Detection of Deepfakes: A Survey](https://mlsanity.com/paper/2004.11138/) (2020)
- [Efficient Training of Audio Transformers with Patchout](https://mlsanity.com/paper/2110.05069/) (2021)
- [Does Audio Deepfake Detection Generalize?](https://mlsanity.com/paper/2203.16263/) (2022)
- [ASVspoof 2021: Towards Spoofed and Deepfake Speech Detection in the Wild](https://mlsanity.com/paper/2210.02437/) (2022)
- [AI-Synthesized Voice Detection Using Neural Vocoder Artifacts](https://mlsanity.com/paper/2304.13085/) (2023)
- [SingFake: Singing Voice Deepfake Detection](https://mlsanity.com/paper/2309.07525/) (2023)
- [FakeMusicCaps: a Dataset for Detection and Attribution of Synthetic Music Generated via Text-to-Music Models](https://mlsanity.com/paper/2409.10684/) (2024)
- [CLAD: Robust Audio Deepfake Detection Against Manipulation Attacks with Contrastive Learning](https://mlsanity.com/paper/2404.15854/) (2024)
- [A Fourier Explanation of AI-music Artifacts](https://mlsanity.com/paper/2506.19108/) (2025)
- [AI-Generated Song Detection via Lyrics Transcripts](https://mlsanity.com/paper/2506.18488/) (2025)

## Abstract (arXiv)

AI-generated music detectors are commonly evaluated against original songs, but real-world uploads are often remixed, re-encoded, pitch-shifted, or otherwise edited. These edited versions form a difficult negative class: they are not generated by AI, yet they may introduce spectral artifacts that resemble synthetic audio fingerprints. We study this problem as a hard-negative robustness setting for AI-generated music detection, focusing on AI-generated and edited variants derived from the same anchor songs. We compile a YouTube-based dataset of AI, edited, and original variants, using the original tracks only as references, and train a binary AI versus edited detector. Audio is processed as 10-second clips and passed as raw waveforms to a pretrained PaSST spectrogram transformer. To reduce leakage, all splits are performed by anchor song. On the held-out test set, the final video-level system achieves 0.811 balanced accuracy. At clip level, AI-generated clips reach an F1-score of 0.836, while edited clips reach a lower F1-score of 0.720. The results suggest that AI-generated music retains detectable fingerprint-like spectral cues beyond ordinary editing, but the lower edited-class F1-score shows that these cues can still overlap with artifacts from edited audio. Grad-CAM visualizations are used to inspect whether high-confidence predictions rely on localized time-frequency regions.