MUSIC PLAGIARISM DETECTION: PROBLEM FORMULATION AND A SEGMENT-BASED SOLUTION

Submitted @ ICCASP 2026

Seonghyeon Go* · Yumin Kim*

MIPPIA Inc.

SMP Dataset Overview

The SMP (Segment-based Music Plagiarism) dataset contains music plagiarism detection pairs with temporal segment annotations. Each row represents a pair of songs with identified similar segments.

Dataset Structure

Column	Description
ori_title	Title of the original song
comp_title	Title of the comparison song
ori_link	YouTube link to the original song
comp_link	YouTube link to the comparison song
relation	Relationship type (plag for plagiarism)
ori_times	List of start times (in seconds) of similar segments in original song
comp_times	List of start times (in seconds) of similar segments in comparison song
pair_number	Unique identifier for song pairs
acoustic_idx	Unique identifier for segment pairs

Data Format

• Time annotations: JSON-formatted lists containing start times of similar segments
• Temporal alignment: ori_times and comp_times correspond to matching similar segments between songs
• Segment duration: Each segment represents a temporally coherent musical phrase or motif

Statistics

• Total pairs: Multiple song pairs with plagiarism relationships
• Temporal annotations: Precise start times for similar musical segments
• Multi-language: Includes both English and Korean songs

License

Our code and demo website are licensed under a GPL License .