v3.2.5

README.md

@@ -315,6 +315,7 @@ The distance measure in parentheses.
 | **DIFGSM**<br />(Linf) | Improving Transferability of Adversarial Examples with Input Diversity ([Xie et al., 2019](https://arxiv.org/abs/1803.06978)) | :heart_eyes: Contributor [taobai](https://github.com/tao-bai) |
 | **TIFGSM**<br />(Linf) | Evading Defenses to Transferable Adversarial Examples by Translation-Invariant Attacks ([Dong et al., 2019](https://arxiv.org/abs/1904.02884)) | :heart_eyes: Contributor [taobai](https://github.com/tao-bai) |
 | **Jitter**<br />(Linf) | Exploring Misclassifications of Robust Neural Networks to Enhance Adversarial Attacks ([Schwinn, Leo, et al., 2021](https://arxiv.org/abs/2105.10304)) | |
+| **Pixle**<br />(L0) | Pixle: a fast and effective black-box attack based on rearranging pixels ([Pomponi, Jary, et al., 2022](https://arxiv.org/abs/2202.02236)) | |
 
 
 

UPDATE_HISTORY.md

@@ -404,4 +404,18 @@
 
 ### v3.2.4
 
-  * `save`, `MultiAttack`: `return_verbose` can be `True `even if `verbose=False`.
+  * `save`, `MultiAttack`: `return_verbose` can be `True `even if `verbose=False`.
+
+
+
+​      
+
+​      
+
+
+### v3.2.5
+
+  * `Pixle` is added.
+  * `save`: Now saving images and labels for every batch.
+  * `OnePixel`: Now supports targeted version.
+  * `_get_target_label`: Now generates target label under evaluation mode and `torch.no_grad()`.

torchattacks/__init__.py

@@ -25,4 +25,4 @@
 from .attacks.jitter import Jitter
 from .attacks.pixle import Pixle
 
-__version__ = '3.2.4'
+__version__ = '3.2.5'