Arbitrary-sized Image Training and Residual Kernel Learning: Towards Image Fraud Identification

TL;DR

This paper introduces a novel framework for training on arbitrarily-sized images without resizing, using pseudo-batch gradient descent and residual kernel learning, significantly improving image fraud detection accuracy.

Contribution

It proposes a new training method for arbitrary-sized images and a residual kernel learning strategy, enhancing image fraud identification performance.

Findings

Achieved state-of-the-art results in image fraud detection.

Effective for images with small tampered regions.

Generalizes well to unseen tampering distributions.

Abstract

Preserving original noise residuals in images are critical to image fraud identification. Since the resizing operation during deep learning will damage the microstructures of image noise residuals, we propose a framework for directly training images of original input scales without resizing. Our arbitrary-sized image training method mainly depends on the pseudo-batch gradient descent (PBGD), which bridges the gap between the input batch and the update batch to assure that model updates can normally run for arbitrary-sized images. In addition, a 3-phase alternate training strategy is designed to learn optimal residual kernels for image fraud identification. With the learnt residual kernels and PBGD, the proposed framework achieved the state-of-the-art results in image fraud identification, especially for images with small tampered regions or unseen images with different tampering distributions.