CVJul 18, 2018

Harmonic Adversarial Attack Method

arXiv:1807.10590v28 citations
Originality Incremental advance
AI Analysis

This addresses the issue of easily spottable adversarial perturbations in computer vision, offering a method to improve stealth and transferability, though it appears incremental by focusing on edge-free perturbations rather than a new paradigm.

The paper tackles the problem of adversarial attacks degrading image quality by proposing Harmonic Adversarial Attack Methods (HAAM), which generate edge-free perturbations using harmonic functions to preserve visual quality even with large magnitudes, and experiments show these adversaries often have higher success rates in transferring between models.

Adversarial attacks find perturbations that can fool models into misclassifying images. Previous works had successes in generating noisy/edge-rich adversarial perturbations, at the cost of degradation of image quality. Such perturbations, even when they are small in scale, are usually easily spottable by human vision. In contrast, we propose Harmonic Adversar- ial Attack Methods (HAAM), that generates edge-free perturbations by using harmonic functions. The property of edge-free guarantees that the generated adversarial images can still preserve visual quality, even when perturbations are of large magnitudes. Experiments also show that adversaries generated by HAAM often have higher rates of success when transferring between models. In addition, we find harmonic perturbations can simulate natural phenomena like natural lighting and shadows. It would then be possible to help find corner cases for given models, as a first step to improving them.

Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes