ART 1.11.1

[beat-buesser]

This release of ART 1.11.1 provides updates to ART 1.11.

Added

[None]

Changed

[None]

Removed

[None]

Fixed

• Fixed unecessary check for targeted attacks in AdversarialPatch and delegated check to framework-specific implementations (Fix untargeted attack in Adversarial Patch #1768)
• Fixed missing transfer to device in AdversarialPatchPyTorch.apply_patch() (Fix bug in apply_patch() function in Adversarial Patch Pytorch running on GPU #1771)
• Fix redundant call to detach().cpu().numpy() in PyTorchClassifier.predict() (Fix exception in pytorch classifier predict method #1785)
• Fixed art.utils.random_sphere() for norm=1 to sample uniformly in the L1 ball (Fix random_sphere() with norm=1 #1802)
• Fixed PyTorch detach() call on Numpy arrays in PyTorchRegressor (Remove call to detach() on a numpy array #1824)
• Fixed probability check for multi-dimensional arrays + out of bounds error in binning in the pointwise differential training privacy metric PDTP (PDTP fixes  #1825)
• Fixed learning rate decay in ElasticNet evasion attack (Fix learning rate decay in ElasticNet attack #1833)

---

This discussion was created from the release ART 1.11.1.