Skip to content

Eenrue/ERPA

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

18 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

README: Robust Error-Aware Digital Signature
via Deep Watermarking Model

Hyunwook Choi1 · Sangyun Won1 · Daeyeon Hwang1 · Junhyeok Choi1

1Korea University

Paper PDF Project Page

🎨 ERPA (ERror PAinting Module) Overview

To enable error-aware watermarking suitable for cryptographic use cases, we propose ERPA, a novel mechanism designed to improve the bit-level reliability of watermark decoding under distortion. ERPA enhances bit-level watermark reliability through a deterministic encoder and a noise-robust neural decoder. The encoder spatially distributes error vectors using a Distinct Circular Subsum Sequence (DCSS), allowing the decoder to recover the original error pattern even under distortion.

📁 Repository Structure

├── model/ # Encoder/Decoder networks

├── scripts/ # Training and evaluation scripts

├── weights/ # Model weights

└── README.md

📌 Citation

@misc{choi2025readmerobusterrorawaredigital,
      title={README: Robust Error-Aware Digital Signature Framework via Deep Watermarking Model}, 
      author={Hyunwook Choi and Sangyun Won and Daeyeon Hwang and Junhyeok Choi},
      year={2025},
      eprint={2507.04495},
      archivePrefix={arXiv},
      primaryClass={cs.CR},
      url={https://arxiv.org/abs/2507.04495}, 
}

About

ERPA (ERror PAinting Module)

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors