Adversarial Color Enhancement: Generating Unrestricted Adversarial Images by Optimizing a Color Filter
This work addresses the challenge of creating unrestricted adversarial modifications for neural networks, moving beyond imperceptible perturbations to focus on perceptible but non-suspicious changes, which is incremental in the context of adversarial machine learning.
The paper tackles the problem of generating adversarial images that fool neural networks by introducing Adversarial Color Enhancement (ACE), which optimizes a color filter via gradient descent to create unrestricted, perceptible yet non-suspicious perturbations, with experimental results validating its white-box adversarial strength and black-box transferability.
We introduce an approach that enhances images using a color filter in order to create adversarial effects, which fool neural networks into misclassification. Our approach, Adversarial Color Enhancement (ACE), generates unrestricted adversarial images by optimizing the color filter via gradient descent. The novelty of ACE is its incorporation of established practice for image enhancement in a transparent manner. Experimental results validate the white-box adversarial strength and black-box transferability of ACE. A range of examples demonstrates the perceptual quality of images that ACE produces. ACE makes an important contribution to recent work that moves beyond $L_p$ imperceptibility and focuses on unrestricted adversarial modifications that yield large perceptible perturbations, but remain non-suspicious, to the human eye. The future potential of filter-based adversaries is also explored in two directions: guiding ACE with common enhancement practices (e.g., Instagram filters) towards specific attractive image styles and adapting ACE to image semantics. Code is available at https://github.com/ZhengyuZhao/ACE.