Research map / Adversarial robustness and security / Attacks on text and graphs
Are You Stealing My Model? Sample Correlation for Fingerprinting Deep Neural Networks
Jiyang Guan, Jian Liang, Ran He
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| SAC-m vs adversarial-example fingerprint CAE against adversarial extraction | SAC · CIFAR10 Extract-Adv | AUC: 0.52 -> 0.92 | Improved |
| SAC-w vs EWE watermark under Finetune-A | SAC · Tiny-ImageNet | AUC: 0.48 -> 1.00 | Improved |
| use CutMix samples instead of normal samples (ablation) | SAC-m · CIFAR10 Extract-Adv | AUC: 0.78 -> 0.92 | Improved |
| use CutMix samples instead of source-misclassified samples (ablation) | SAC-m · CIFAR10 Extract-L | AUC: 0.88 -> 1.00 | Improved |
| cosine similarity instead of Gaussian RBF as correlation function (ablation) | SAC · CIFAR10 | AUC: 1.0 -> 1.0 | No clear effect |
| predicted-label-only access instead of probabilities (ablation) | SAC · CIFAR10 Transfer-A | AUC: 1.0 -> 0.0 | Hurt |
| predicted-label-only access instead of probabilities (ablation) | SAC · CIFAR10 Transfer-L | AUC: 1.0 -> 0.9 | No clear effect |
| SAC-m vs CAE fingerprint generation | SAC-m · CIFAR10 time | seconds: 25,536.89s -> 4.45s | Improved |
| EWE watermarking vs fingerprinting on source accuracy | EWE · CIFAR10 source accuracy | accuracy: 87.3 -> 83.3(-4.0) | Hurt |
Closest papers by meaning
- Sample Correlation for Fingerprinting Deep Face Recognition (IJCV 2024 · similarity 0.94)
- Queries, Representation & Detection: The Next 100 Model Fingerprinting Schemes (AAAI 2025 · similarity 0.87)
- Fully Exploiting Every Real Sample: SuperPixel Sample Gradient Model Stealing (CVPR 2024 · similarity 0.85)
- Stealix: Model Stealing via Prompt Evolution (ICML 2025 · similarity 0.84)
- Efficient Model Stealing Defense with Noise Transition Matrix (CVPR 2024 · similarity 0.84)
- United We Stand, Divided We Fall: Fingerprinting Deep Neural Networks via Adversarial Trajectories (NeurIPS 2024 · similarity 0.84)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
