Training CNNs in Presence of JPEG Compression: Multimedia Forensics vs Computer Vision

TL;DR

This paper investigates how JPEG compression impacts CNN training for both forensic and computer vision tasks, emphasizing the importance of considering compression effects for forensic applications to maintain generalization.

Contribution

It highlights the different effects of JPEG compression on CNN training for forensic versus computer vision tasks and provides guidelines for dataset generation.

Findings

JPEG compression significantly affects forensic CNN training.

For computer vision tasks, JPEG effects can often be ignored.

Proper dataset generation considering JPEG effects improves forensic detector performance.

Abstract

Convolutional Neural Networks (CNNs) have proved very accurate in multiple computer vision image classification tasks that required visual inspection in the past (e.g., object recognition, face detection, etc.). Motivated by these astonishing results, researchers have also started using CNNs to cope with image forensic problems (e.g., camera model identification, tampering detection, etc.). However, in computer vision, image classification methods typically rely on visual cues easily detectable by human eyes. Conversely, forensic solutions rely on almost invisible traces that are often very subtle and lie in the fine details of the image under analysis. For this reason, training a CNN to solve a forensic task requires some special care, as common processing operations (e.g., resampling, compression, etc.) can strongly hinder forensic traces. In this work, we focus on the effect that JPEG has on CNN training considering different computer vision and forensic image classification problems. Specifically, we consider the issues that rise from JPEG compression and misalignment of the JPEG grid. We show that it is necessary to consider these effects when generating a training dataset in order to properly train a forensic detector not losing generalization capability, whereas it is almost possible to ignore these effects for computer vision tasks.