Enhancing Neural Audio Fingerprint Robustness to Audio Degradation for Music Identification

Audio fingerprinting (AFP) allows the identification of unknown audio content by extracting compact representations, termed audio fingerprints, that are designed to remain robust against common audio degradations. Neural AFP methods often employ metric learning, where representation quality is influenced by the nature of the supervision and the utilized loss function. However, recent work unrealistically simulates real-life audio degradation during training, yielding sub-optimal supervision. Additionally, although several modern metric learning approaches have been proposed, current neural AFP methods continue to rely on the NT‑Xent loss without exploring the recent advances or classical alternatives. In this work, we propose a series of best practices to enhance self-supervision by leveraging musical signal properties and realistic room acoustics. We then present the first systematic evaluation of various metric learning approaches in the context of AFP, demonstrating that a self‑supervised adaptation of the triplet loss yields superior performance. Our results also reveal that training with multiple positives per anchor has critically different effects across loss functions. Our approach, termed NMFP, is built upon these insights and achieves state-of-the-art performance on both a large, synthetically degraded dataset and an industrial dataset recorded using real microphones in diverse music venues.