Research map / Adversarial robustness and security / Black-box and transfer attacks
Boosting the Uniqueness of Neural Networks Fingerprints with Informative Triggers
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's abstract only. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Proposes informative fingerprinting triggers to improve the uniqueness of black-box neural network fingerprints for copyright tracing.
Problem
Black-box DNN copyright tracing via fingerprints (outputs on fingerprinting triggers) is usually evaluated for robustness, while the accuracy and uniqueness of copyright tracing (distinguishing independently trained models) is overlooked.
Approach
Designs fingerprinting triggers that are informative, so that fingerprints of different networks are more distinguishable (uniqueness), rather than only optimizing robustness. Based on the abstract only; details not verified.
Closest papers by meaning
- Revisiting the Information Capacity of Neural Network Watermarks: Upper Bound Estimation and Beyond (AAAI 2024 · similarity 0.86)
- SDBF: Steep-Decision-Boundary Fingerprinting for Hard-Label Tampering Detection of DNN Models (CVPR 2025 · similarity 0.86)
- IPRemover: A Generative Model Inversion Attack against Deep Neural Network Fingerprinting and Watermarking (AAAI 2024 · similarity 0.85)
- United We Stand, Divided We Fall: Fingerprinting Deep Neural Networks via Adversarial Trajectories (NeurIPS 2024 · similarity 0.85)
- Intersecting-Boundary-Sensitive Fingerprinting for Tampering Detection of DNN Models (ICML 2024 · similarity 0.84)
- Resource Efficient Deep Learning Hardware Watermarks with Signature Alignment (AAAI 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.