Abstract:Using images that are dispersed in a network can improve image classification performance; however, it is easy to leak the image holders' privacy. In this paper, we present a novel method that considers both image classification performance and the image holders' privacy-preservation. Our method contains four stages; the feature extraction stage follows the Bag-of-Features (BoF) framework, but it introduces significantly novel stages in the training, classification and ensemble stages. In the training stage, we split the image classification into many one-vs.-one classification problems; a set of binary classifiers as weak learners are learned on a local machine, to classify each image in an autonomous system. In the classification stage, a local machine sends features of unlabeled images to each image autonomous system, and then, each image autonomous system and local machine finish a prediction task under their own trained model. Afterward, the image autonomous systems send their prediction labels back to the local system. In the ensemble stage, we gather all of the prediction labels from the image autonomous systems and design an ensemble system to ensemble them into final classification results. This ensemble system must consider both the efficiency of the different trained models in each image autonomous system and the efficiency of the trained model on each class in the one-vs.-one classification problems. During the whole process, only the features of unlabeled images and prediction labels are transmitted in the network, which largely reduces the risk of privacy leakage. We conducted experiments on the network environment, to prove that our method has preferable performance. Our first experiment demonstrates that the performance of our method achieves state-of-the-art performance on four public dataset benchmarks, which shows that our method has competitive performance. The second experiment shows that our method achieves state-of-the-art performance with even fewer training samples, which means that there is less communication overhead and better privacy-preservation.

Graphic image classification method based on an attention mechanism and fusion of multilevel and multiscale deep features

Learning Multi-Level and Multi-Scale Deep Representations for Privacy Image Classification

Hyperspectral and Multispectral Image Fusion Based on Deep Attention Network.

3D Point Cloud Classification Method Based on Multiple Attention Mechanism and Dynamic Graph Convolution

Dynamic Deep Multi-modal Fusion for Image Privacy Prediction

An image classification method that considers privacy-preservation.

Deep Multi-Level Fusion Network for Multi-Source Image Pixel-Wise Classification

Image Sentiment Analysis Method Based on Multi-Level Feature Fusion

When graph convolution meets double attention: online privacy disclosure detection with multi-label text classification

Aerial Scene Classification Via Multilevel Fusion Based on Deep Convolutional Neural Networks.

Novel Cross-Resolution Feature-Level Fusion for Joint Classification of Multispectral and Panchromatic Remote Sensing Images

A Privacy-Preserving Image Retrieval Method Based on Deep Learning and Adaptive Weighted Fusion.

GAFnet: Group Attention Fusion Network for PAN and MS Image High-Resolution Classification

Research on Image Semantic Segmentation Based on Hybrid Cascade Feature Fusion and Detailed Attention Mechanism

Image Tampering Detection Method Based on Multi-Feature Fusion

Combining Spatial Attention and Cross-Layer Bilinear Pooling for Fine-Grained Image Classification

A Lightweight Privacy-Preserving Scheme Using Label-based Pixel Block Mixing for Image Classification in Deep Learning

Image Privacy Prediction Using Deep Neural Networks

Hiding Private Information in Images from AI.

A Multi-Feature Fusion Model Based on Denoising Convolutional Neural Network and Attention Mechanism for Image Classification

A Multiscale Dual-Branch Feature Fusion and Attention Network for Hyperspectral Images Classification