# Compression
python compress.py <file> <file>.ms
# Decompression
python decompress.py <file>.ms <file>.ms.out
| Dataset | Type | Size (B) | Link |
|---|---|---|---|
| SAO | homogeneous | 7251944 | https://sun.aei.polsl.pl/~sdeor/corpus/silesia.zip |
| NCI | homogeneous | 33553445 | https://sun.aei.polsl.pl/~sdeor/corpus/silesia.zip |
| X-RAY | image | 8474240 | https://sun.aei.polsl.pl/~sdeor/corpus/silesia.zip |
| MR | image | 9970564 | https://sun.aei.polsl.pl/~sdeor/corpus/silesia.zip |
| HaHi | genome | 3890005 | https://sweet.ua.pt/pratas/datasets/DNACorpus.zip |
| DrMe | genome | 32181429 | https://sweet.ua.pt/pratas/datasets/DNACorpus.zip |
If you are interested in our work, we hope you might consider starring our repository and citing our paper:
@inproceedings{ma2025multi,
title={Multi-source Data Lossless Compression via Parallel Expansion Mapping and xLSTM},
author={Ma, Huidong and Sun, Hui and Yi, Liping and Liu, Xiaoguang and Wang, Gang},
booktitle={ICASSP 2025-2025 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)},
pages={1--5},
year={2025},
organization={IEEE}
}
The code is based on xLSTM and Reference-arithmetic-coding. Thanks for these great works.
Email: mahd@nbjl.nankai.edu.cn
Nankai-Baidu Joint Laboratory (NBJL)