# Supervised contrastive learning from weakly-labeled audio segments for musical version matching

arXiv: https://arxiv.org/abs/2502.16936 · HTML: https://mlsanity.com/paper/2502.16936/

## Key facts

- Cited by 8 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 24 Feb 2025.

## Built on by (most cited first, 8 of 8)

- [Attribution-by-design: Ensuring Inference-Time Provenance in Generative Music Systems](https://mlsanity.com/paper/2510.08062/) (2025)
- [Leveraging Whisper Embeddings for Audio-based Lyrics Matching](https://mlsanity.com/paper/2510.08176/) (2025)
- [Automatic Music Sample Identification with Multi-Track Contrastive Learning](https://mlsanity.com/paper/2510.11507/) (2025)
- [Scalable Music Cover Retrieval Using Lyrics-Aligned Audio Embeddings](https://mlsanity.com/paper/2601.11262/) (2026)
- [Yes, But Not Always. Generative AI Needs Nuanced Opt-in](https://mlsanity.com/paper/2604.09413/) (2026)
- [Rubato: Transcribing Piano Music with Timestamps](https://mlsanity.com/paper/2605.24291/) (2026)
- [YuE2: Unifying Symbolic and Audio Music Generation at Frontier Quality](https://mlsanity.com/paper/2609.33757/) (2026)
- [Ghost in the Encoder: Decodable Artist Identity Representations in Lyrics-to-Song Generation](https://mlsanity.com/paper/2609.39552/) (2026)

## Abstract (arXiv)

Detecting musical versions (different renditions of the same piece) is a challenging task with important applications. Because of the ground truth nature, existing approaches match musical versions at the track level (e.g., whole song). However, most applications require to match them at the segment level (e.g., 20s chunks). In addition, existing approaches resort to classification and triplet losses, disregarding more recent losses that could bring meaningful improvements. In this paper, we propose a method to learn from weakly annotated segments, together with a contrastive loss variant that outperforms well-studied alternatives. The former is based on pairwise segment distance reductions, while the latter modifies an existing loss following decoupling, hyper-parameter, and geometric considerations. With these two elements, we do not only achieve state-of-the-art results in the standard track-level evaluation, but we also obtain a breakthrough performance in a segment-level evaluation. We believe that, due to the generality of the challenges addressed here, the proposed methods may find utility in domains beyond audio or musical version matching.