How can i use the pretrained model

[slasz]

Do you have any sample code to use pre-trained model to crop a single image input, or any suggestion ?

[bo-zhang-cs]

Please refer to the test.py. To perform prediction on single image, you need replace the dataloader of test.py with your image file and then preprocess image in the same manner as in Cropping_dataset.py before sending to the network.

[slasz]

Please refer to the test.py. To perform prediction on single image, you need replace the dataloader of test.py with your image file and then preprocess image in the same manner as in Cropping_dataset.py before sending to the network.

Thank you :)

[slasz]

def test_pred(model):
    print('Call test predict !!!')
    img_test = 'img/mm.jpg'
    im = cv2.imread(img_test)
    image_transformer = transforms.Compose([
        transforms.ToTensor(),
        transforms.Normalize(mean=IMAGE_NET_MEAN, std=IMAGE_NET_STD)])
    im_ts = image_transformer(im)
    logits, kcm, crop = model(im_ts, only_classify=False)

if __name__ == '__main__':
    weight_file = "./pretrained_model/best-FLMS_iou.pth"
    model = CACNet(loadweights=True)
    model.load_state_dict(torch.load(weight_file,map_location=device))
    model = model.to(device).eval()
    test_pred(model)

---

I try to make the function and call from main and got error below

RuntimeError: Input type (torch.FloatTensor) and weight type (torch.cuda.FloatTensor) should be the same or input should be a MKLDNN tensor and weight is a dense tensor

[slasz]

Now I can run by making class SingleDataset that inherits from Dataset and I have more questions to ask about can I crop with specific image size

[slasz]

I can run with single image already