Pic should be Tensor or ndarray. Got <class ‘dict’>

vision
1

**I tried in previous discussions but I need to transform both the keypoint values ​​and the photo in same size

like the tutorial, I cannot extract the data I have given as a sample-dict to examine it after I have converted it and I cannot feed it into the model. Where exactly is the problem?

My transform code :
transform = transforms.Compose([transforms.ToPILImage(),transforms.Resize(256),transforms.RandomCrop(224),transforms.ToTensor(),transforms.Normalize(mean=[0.485, 0.456, 0.406],std=[0.229, 0.224, 0.225])])

and

transformedDataset = keypointData(csv_path=keypoint_path_training,image_path=image_train,transform=transform)
print('Number of transformed images: ', len(transformedDataset))
valid = int(2*len(transformedDataset)/10)
train_set, val_set = torch.utils.data.random_split(transformedDataset, [len(transformedDataset)-valid, valid])
train_set = torch.utils.data.DataLoader(train_set,batch_size = 16,shuffle = True,num_workers=0)
val_set = torch.utils.data.DataLoader(val_set,batch_size = 16,shuffle = True,num_workers=0)

inputsTrain= next(iter(train_set))
inputsValid = next(iter(val_set))

pic should be Tensor or ndarray. Got <class ‘dict’>.

i try to convert numpy array but its not running

2

I’m not sure, if you are following this tutorial, but if so, note that the transformations are taken from references/detection/transforms.py as described in the tutorial, so you might need to use them as well.
The difference to torchvision.transforms is that these transformation accept the image and target tensors.