Backdoor-free model watermarking via differential verification of adversarial samples and multi-bit information extraction
Abstract Model watermarking is a commonly used ownership verification technique for protecting the copyright of deep learning models. However, in practically deployed black-box model service scenarios, existing methods typically rely on misclassification-based backdoor trigger mechanisms, or assume...
Gorde:
| Egile Nagusiak: | , , , , |
|---|---|
| Formatua: | Artigo |
| Hizkuntza: | Inglês |
| Argitaratua: |
Springer
2026-08-01
|
| Saila: | Journal of King Saud University: Computer and Information Sciences |
| Gaiak: | |
| Sarrera elektronikoa: | https://doi.org/10.1007/s44443-026-01195-3 |
| Etiketak: |
Etiketarik gabe, Izan zaitez lehena erregistro honi etiketa jartzen!
|