ART 1.10.1

[beat-buesser]

This release of ART 1.10.1 provides updates to ART 1.10.

Added

[None]

Changed

• Changed AdversarialTrainerMadryPGD.fit to support arguments nb_epochs and batch_size (Add nb_epochs and batch_size to AdversarialTrainerMadryPGD.fit #1612)
• Changed GradientMatchingAttack to add support for models with undefined input shape by abstracting the shape information from the input data (Gradient matching unknown input shape support #1624)
• Changed PyTorchObjectDetector to support inputs with number of channels other than 1 and 3 (Add support for multi-channel images in PyTorchObjectDetector #1633)

Removed

[None]

Fixed

• Fixed incorrect handling of true regression labels in attribute inference attacks (Correct handling of labels for regression models in attribute attacks #1598)
• Fixed AdversarialPatchPyTorch.apply_patch to correctly check if mask is None (Check if mask is None in AdversarialPatchPyTorch.apply_patch #1607)

---

This discussion was created from the release ART 1.10.1.