Datasets:
File size: 2,235 Bytes
106f89b 465c30b 106f89b 465c30b | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 | ---
license: mit
language:
- en
- ko
tags:
- music
- music-information-retrieval
- plagiarism-detection
- music-similarity
task_categories:
- audio-classification
size_categories:
- n<1K
configs:
- config_name: default
data_files:
- split: train
path: smp_dataset.csv
arxiv: 2601.21260
---
# Similar Music Pair (SMP) Dataset
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.
This dataset accompanies the paper
[Music Plagiarism Detection: Problem Formulation and a Segment-based Solution](https://arxiv.org/abs/2601.21260).
- Code: https://github.com/Mippia/Music-Plagiarism-Detection
- Demo: https://huggingface.co/spaces/mippia/MPD-demo
- Project page: https://mippia.github.io/icassp-mpd/
## 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 (e.g., `plag` for plagiarism) |
| `ori_times` | List of start times (seconds) of similar segments in the original song |
| `comp_times` | List of start times (seconds) of similar segments in the comparison song |
| `pair_number` | Unique identifier for song pairs |
| `acoustic_idx` | Unique identifier for segment pairs |
Time annotations are JSON-formatted lists; `ori_times[i]` and `comp_times[i]`
correspond to the same matching segment between the two songs.
## Usage
```python
from datasets import load_dataset
dataset = load_dataset("mippia/SMP-dataset")
print(dataset["train"][0])
```
## Notes
- **Audio is not included.** Only YouTube links and temporal annotations are
provided. Users must retrieve audio from YouTube themselves.
- The dataset includes both English and Korean songs across multiple genres.
## Citation
```bibtex
@inproceedings{go2026mpd,
title={Music Plagiarism Detection: Problem Formulation and a Segment-based Solution},
author={Go, Seonghyeon and Kim, Yumin},
booktitle={ICASSP},
year={2026}
}
```
## License
Released under MIT License. |