Beyond Pixel Norm-Balls: Parametric Adversaries using an Analytically Differentiable Renderer
This addresses the vulnerability of image classifiers to more realistic adversarial attacks, representing an incremental advance in adversarial robustness evaluation.
The paper tackles the limited practical utility of pixel-based adversarial attacks by proposing parametric norm-balls that perturb physical parameters like lighting and geometry, using a differentiable renderer to enable physically-based adversarial attacks with scalable data augmentation.
Many machine learning image classifiers are vulnerable to adversarial attacks, inputs with perturbations designed to intentionally trigger misclassification. Current adversarial methods directly alter pixel colors and evaluate against pixel norm-balls: pixel perturbations smaller than a specified magnitude, according to a measurement norm. This evaluation, however, has limited practical utility since perturbations in the pixel space do not correspond to underlying real-world phenomena of image formation that lead to them and has no security motivation attached. Pixels in natural images are measurements of light that has interacted with the geometry of a physical scene. As such, we propose the direct perturbation of physical parameters that underly image formation: lighting and geometry. As such, we propose a novel evaluation measure, parametric norm-balls, by directly perturbing physical parameters that underly image formation. One enabling contribution we present is a physically-based differentiable renderer that allows us to propagate pixel gradients to the parametric space of lighting and geometry. Our approach enables physically-based adversarial attacks, and our differentiable renderer leverages models from the interactive rendering literature to balance the performance and accuracy trade-offs necessary for a memory-efficient and scalable adversarial data augmentation workflow.