8f420a3b35944dcfa470dee958ef61dee221ce02,implementations/dualgan/datasets.py,MNISTM,__getitem__,#MNISTM#Any#,60

Before Change



        // doing this so that it is consistent with all other datasets
        // to return a PIL Image
        img = Image.fromarray(img.squeeze().numpy(), mode="RGB")

        if self.transform is not None:
            img = self.transform(img)

After Change



        img_pair = self.transform(Image.open(self.files[index % len(self.files)]))
        _, h, w = img_pair.shape
        half_w = int(w/2)

        item_A = img_pair[:, :, :half_w]
        item_B = img_pair[:, :, half_w:]
Italian Trulli
In pattern: SUPERPATTERN

Frequency: 3

Non-data size: 4

Instances


Project Name: eriklindernoren/PyTorch-GAN
Commit Name: 8f420a3b35944dcfa470dee958ef61dee221ce02
Time: 2018-04-23
Author: eriklindernoren@gmail.com
File Name: implementations/dualgan/datasets.py
Class Name: MNISTM
Method Name: __getitem__


Project Name: IBM/adversarial-robustness-toolbox
Commit Name: e21ef336207b0f9ae378c77430d298945827830a
Time: 2019-02-12
Author: M.N.Tran@ibm.com
File Name: art/classifiers/pytorch.py
Class Name: PyTorchClassifier
Method Name: get_activations


Project Name: IBM/adversarial-robustness-toolbox
Commit Name: a922855fe7aef360e4a360caa800388d9c843355
Time: 2019-07-23
Author: beat.buesser@ie.ibm.com
File Name: art/classifiers/tensorflow.py
Class Name: TensorflowV2Classifier
Method Name: predict