# Echoes: A semantically-aligned music deepfake detection dataset

arXiv: https://arxiv.org/abs/2603.23667 · HTML: https://mlsanity.com/paper/2603.23667/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 27 of its 34 references that are in the map.
- First posted on arXiv 24 Mar 2026.

## Builds on (most cited first, 27 of 27)

- [wav2vec 2.0: A Framework for Self-Supervised Learning of Speech Representations](https://mlsanity.com/paper/2006.11477/) (2020)
- [XLS-R: Self-supervised Cross-lingual Speech Representation Learning at Scale](https://mlsanity.com/paper/2111.09296/) (2021)
- [AudioLDM: Text-to-Audio Generation with Latent Diffusion Models](https://mlsanity.com/paper/2301.12503/) (2023)
- [Simple and Controllable Music Generation](https://mlsanity.com/paper/2306.05284/) (2023)
- [MusicLM: Generating Music From Text](https://mlsanity.com/paper/2301.11325/) (2023)
- [FMA: A Dataset For Music Analysis](https://mlsanity.com/paper/1612.01840/) (2016)
- [Stable Audio Open](https://mlsanity.com/paper/2407.14358/) (2024)
- [AASIST: Audio Anti-Spoofing using Integrated Spectro-Temporal Graph Attention Networks](https://mlsanity.com/paper/2110.01200/) (2021)
- [ASVspoof 2019: Future Horizons in Spoofed and Fake Audio Detection](https://mlsanity.com/paper/1904.05441/) (2019)
- [End-to-end anti-spoofing with RawNet2](https://mlsanity.com/paper/2011.01108/) (2020)
- [Does Audio Deepfake Detection Generalize?](https://mlsanity.com/paper/2203.16263/) (2022)
- [Automatic speaker verification spoofing and deepfake detection using wav2vec 2.0 and data augmentation](https://mlsanity.com/paper/2202.12233/) (2022)
- [ACE-Step: A Step Towards Music Generation Foundation Model](https://mlsanity.com/paper/2506.00045/) (2025)
- [DiffRhythm: Blazingly Fast and Embarrassingly Simple End-to-End Full-Length Song Generation with Latent Diffusion](https://mlsanity.com/paper/2503.01183/) (2025)
- [Speech is Silver, Silence is Golden: What do ASVspoof-trained Models Really Learn?](https://mlsanity.com/paper/2106.12914/) (2021)
- [SongGen: A Single Stage Auto-regressive Transformer for Text-to-Song Generation](https://mlsanity.com/paper/2502.13128/) (2025)
- [Towards generalisable and calibrated synthetic speech detection with self-supervised representations](https://mlsanity.com/paper/2309.05384/) (2023)
- [Circumventing shortcuts in audio-visual deepfake detection datasets with unsupervised learning](https://mlsanity.com/paper/2412.00175/) (2024)
- [SingFake: Singing Voice Deepfake Detection](https://mlsanity.com/paper/2309.07525/) (2023)
- [SONICS: Synthetic Or Not -- Identifying Counterfeit Songs](https://mlsanity.com/paper/2408.14080/) (2024)
- [Spoofed training data for speech spoofing countermeasure can be efficiently created using neural vocoders](https://mlsanity.com/paper/2210.10570/) (2022)
- [Benchmarking Music Generation Models and Metrics via Human Preference Studies](https://mlsanity.com/paper/2506.19085/) (2025)
- [FakeMusicCaps: a Dataset for Detection and Attribution of Synthetic Music Generated via Text-to-Music Models](https://mlsanity.com/paper/2409.10684/) (2024)
- [From Audio Deepfake Detection to AI-Generated Music Detection -- A Pathway and Overview](https://mlsanity.com/paper/2412.00571/) (2024)
- [Singing Voice Graph Modeling for SingFake Detection](https://mlsanity.com/paper/2406.03111/) (2024)
- [SVDD 2024: The Inaugural Singing Voice Deepfake Detection Challenge](https://mlsanity.com/paper/2408.16132/) (2024)
- [Detecting music deepfakes is easy but actually hard](https://mlsanity.com/paper/2405.04181/) (2024)

## Abstract (arXiv)

We introduce Echoes, a new dataset for music deepfake detection designed for training and benchmarking detectors under realistic and provider-diverse conditions. Echoes comprises 4,468 tracks (131 hours of audio) spanning multiple genres (pop, rock, electronic), and includes content generated by ten popular AI music generation systems. To prevent shortcut learning and promote robust generalization, the dataset is deliberately constructed to be challenging, enforcing semantic-level alignment between spoofed audio and bona fide references. This alignment is achieved by conditioning generated audio samples directly on bona-fide waveforms or song descriptors. We evaluate Echoes in a cross-dataset setting against three existing AI-generated music datasets using state-of-the-art Wav2Vec2 XLS-R 2B representations. Results show that (i) Echoes is the hardest in-domain dataset; (ii) detectors trained on existing datasets transfer poorly to Echoes; (iii) training on Echoes yields the strongest generalization performance. These findings suggest that provider diversity and semantic alignment help learn more transferable detection cues.