Lightweight CNN Model Hashing with Higher-Order Statistics and Chaotic Mapping for Piracy Detection and Tamper Localization

TL;DR

A lightweight CNN model hashing technique that integrates higher-order statistics features with a chaotic mapping mechanism that enables efficient piracy detection and precise tampering localization and is validated for model copyright protection and integrity verification.

Abstract

With the widespread adoption of deep neural networks (DNNs), protecting intellectual property and detecting unauthorized tampering of models have become pressing challenges. Recently, Perceptual hashing has emerged as an effective approach for identifying pirated models. However, existing methods either rely on neural networks for feature extraction, demanding substantial training resources, or suffer from limited applicability and cannot be universally applied to all convolutional neural networks (CNNs). To address these limitations, we propose a lightweight CNN model hashing technique that integrates higher-order statistics (HOS) features with a chaotic mapping mechanism. Without requiring any auxiliary neural network training, our method enables efficient piracy detection and precise tampering localization. Specifically, we extract skewness, kurtosis, and structural features from the parameters of each network layer to construct a model hash that is both robust and discriminative. Additionally, we introduce chaotic mapping to amplify minor changes in model parameters by exploiting the sensitivity of chaotic systems to initial conditions, thereby facilitating accurate localization of tampered regions. Experimental results validate the effectiveness and practical value of the proposed method for model copyright protection and integrity verification.

Figures (6)

• Figure 1: An illustration of limitations inherent to NTS-based method.
• Figure 2: Comparison of parameter distributions between ResNet32 models trained with L2 regularization and Kaiming initialization versus uniform initialization without regularization.
• Figure 3: For the two-dimensional system, the bifurcation diagrams of the LEs (top) and the state variables $x$ and $q$ (bottom) as functions of the parameter $k$.
• Figure 4: Normalized hamming distances of hash codes between the original models and the fine-tuned versions using NTS and HOS methods.
• Figure 5: Normalized hamming distances of hash codes between the original models and the pruned versions using NTS and HOS methods.