Music Plagiarism Detection: Problem Formulation and a Segment-based Solution

TL;DR

This work defines Music Plagiarism Detection as a task distinct from existing MIR problems by requiring precise segment-level localization of plagiarized material ($A'$) within a database for a query track ($A$). It proposes a segment-transcription pipeline that converts audio into structured representations and employs both music-domain similarity and deep learning–based similarity within a multimodal Siamese framework to detect and explain segment-level plagiarism. A new SMP dataset with time-annotated plagiarized segments, along with experiments on 4-bar segments, demonstrates the feasibility and challenges of segment-level plagiarism detection, revealing strengths in segment-level explainability and limitations in matching full-song similarity. The study highlights the need for tailored benchmarks, metrics, and models optimized for partial, element-specific plagiarism, with practical implications for copyright analysis and forensics in music ecosystems.

Abstract

Recently, the problem of music plagiarism has emerged as an even more pressing social issue. As music information retrieval research advances, there is a growing effort to address issues related to music plagiarism. However, many studies, including our previous work, have conducted research without clearly defining what the music plagiarism detection task actually involves. This lack of a clear definition has slowed research progress and made it hard to apply results to real-world scenarios. To fix this situation, we defined how Music Plagiarism Detection is different from other MIR tasks and explained what problems need to be solved. We introduce the Similar Music Pair dataset to support this newly defined task. In addition, we propose a method based on segment transcription as one way to solve the task. Our demo and dataset are available at https://github.com/Mippia/ICASSP2026-MPD.

Figures (1)

• Figure 1: Overall architecture of the proposed music plagiarism detection system. The pipeline processes input audio through segment transcription, extracts musical features for each instrument and structural component, and applies both algorithmic and deep learning-based similarity analysis to detect plagiarism cases.