Adversarial Color Enhancement: Generating Unrestricted Adversarial Images by Optimizing a Color Filter
arXiv:2002.01008
Abstract
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 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.
arXiv commentAccepted by BMVC 2020. Code is available at https://github.com/ZhengyuZhao/ACE
References in corpus (12)
- Very Deep Convolutional Networks for Large-Scale Image Recognition
- Deep Bilateral Learning for Real-Time Image Enhancement
- Exploring the Landscape of Spatial Robustness
- Spatially Transformed Adversarial Examples
- A General Framework for Adversarial Examples with Objectives
- Motivating the Rules of the Game for Adversarial Example Research
- Adversarial Examples for Semantic Segmentation and Object Detection
- Functional Adversarial Attacks
- Unrestricted Adversarial Examples
- Unrestricted Adversarial Examples via Semantic Manipulation
- Panoptic Feature Pyramid Networks
- Wasserstein Adversarial Examples via Projected Sinkhorn Iterations