All You Need is RAW: Defending Against Adversarial Attacks with Camera Image Pipelines

TL;DR

This paper introduces a RAW-based preprocessing defense for neural networks that leverages the natural image acquisition process to effectively mitigate adversarial attacks across multiple vision tasks without retraining.

Contribution

It proposes a model-agnostic RAW-to-RGB pipeline that enhances adversarial robustness by exploiting the natural image capture process, without requiring adversarial training.

Findings

Outperforms existing defenses on ImageNet and COCO datasets

Effective across classification, segmentation, and detection tasks

Does not require retraining for new tasks

Abstract

Existing neural networks for computer vision tasks are vulnerable to adversarial attacks: adding imperceptible perturbations to the input images can fool these methods to make a false prediction on an image that was correctly predicted without the perturbation. Various defense methods have proposed image-to-image mapping methods, either including these perturbations in the training process or removing them in a preprocessing denoising step. In doing so, existing methods often ignore that the natural RGB images in today's datasets are not captured but, in fact, recovered from RAW color filter array captures that are subject to various degradations in the capture. In this work, we exploit this RAW data distribution as an empirical prior for adversarial defense. Specifically, we proposed a model-agnostic adversarial defensive method, which maps the input RGB images to Bayer RAW space and back to output RGB using a learned camera image signal processing (ISP) pipeline to eliminate potential adversarial patterns. The proposed method acts as an off-the-shelf preprocessing module and, unlike model-specific adversarial training methods, does not require adversarial images to train. As a result, the method generalizes to unseen tasks without additional retraining. Experiments on large-scale datasets (e.g., ImageNet, COCO) for different vision tasks (e.g., classification, semantic segmentation, object detection) validate that the method significantly outperforms existing methods across task domains.