Closed Janspiry closed 2 years ago
Hi,thanks for this great works. I used the pretrained model to test on Standford Car dataset, but get the same results in most images. I used the transform function on below:
transforms.Resize(int(368 / 1)), transforms.CenterCrop(368), transforms.ToTensor()
and read the image by:
img = Image.open(img_path).convert("RGB") img = self.tfs(img)
Is there anything I need to pay extra attention to?
the "transforms.CenterCrop(368)" is wrong, remove it (squish resizing)
other than that, triple check to make sure you are loading the model correctly
Hi,thanks for this great works. I used the pretrained model to test on Standford Car dataset, but get the same results in most images. I used the transform function on below:
and read the image by:
Is there anything I need to pay extra attention to?