MelodySim: Measuring Melody-aware Music Similarity for Plagiarism Detection

2025-05-27

AI Summary

This work addresses melody-aware music plagiarism detection. To this end, we introduce MelodySim—the first benchmark dataset specifically designed for melody similarity assessment—and propose a melody-preserving MIDI augmentation paradigm (e.g., note splitting, arpeggiation, non-bass track dropout) to generate high-fidelity variants that enhance model generalization. Methodologically, we develop a segment-level melody similarity discrimination model that integrates the MERT encoder with a triplet network, yielding an interpretable, fine-grained plagiarism localization decision matrix. Our approach achieves high accuracy on MelodySim, and user studies confirm its cognitive validity in melody similarity judgment. Key contributions include: (1) the first melody-specific benchmark dataset; (2) the first melody-aware MIDI data augmentation framework; and (3) an explainable, segment-level plagiarism localization architecture grounded in perceptual melody representation.

Technology Category

• Machine Learning: Multimodal Learning
• Cognitive Modeling & Cognitive Systems: Computational Creativity
• Data Mining & Knowledge Management: Mining of Visual, Multimedia & Multimodal Data

Application Category

• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems
• Web Mining and Content Analysis: Web data generation and simulation
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

We propose MelodySim, a melody-aware music similarity model and dataset for plagiarism detection. First, we introduce a novel method to construct a dataset with focus on melodic similarity. By augmenting Slakh2100; an existing MIDI dataset, we generate variations of each piece while preserving the melody through modifications such as note splitting, arpeggiation, minor track dropout (excluding bass), and re-instrumentation. A user study confirms that positive pairs indeed contain similar melodies, with other musical tracks significantly changed. Second, we develop a segment-wise melodic-similarity detection model that uses a MERT encoder and applies a triplet neural network to capture melodic similarity. The resultant decision matrix highlights where plagiarism might occur. Our model achieves high accuracy on the MelodySim test set.

Problem

• Measuring melody-aware music similarity for plagiarism detection
• Constructing dataset with melodic similarity through MIDI augmentation
• Developing segment-wise melodic-similarity detection model using MERT encoder

Innovation

• Augmented Slakh2100 dataset with melody-preserving variations
• Segment-wise similarity model using MERT encoder
• Triplet neural network for melodic similarity detection