Research Article
AEGuard: Image Feature-Based Independent Adversarial Example Detection Model
Table 2
Proposed system training dataset attributes.
| Attribute | Datatype | Correlation | Explanation |
| edge_noise | Double | 0.6974 | Ratio of noise to the edge pixels | edge_density | Double | 0.5450 | Ratio of edge pixels to the total pixels of image | dct_bias | Double | ā0.8172 | DCT bias of the image | color_composition | Double (3) | 0.1193 | Percentage of each color channel (R, G, and B) of the image | Variance | Double | 0.2620 | Image variance of the image | Entropy | Double | 0.1522 | Image entropy of the image | is_adv | Boolean | ā | Normal image/adversarial example |
|
|