byungjae89 / SPADE-pytorch

PyTorch implementation of "Sub-Image Anomaly Detection with Deep Pyramid Correspondences"
Apache License 2.0
237 stars 43 forks source link

ValueError:continuous format is not supported #5

Closed renqiqi closed 3 years ago

renqiqi commented 3 years ago

Hi, When I trained the model on my own dataset, there is a problem: ValueError: continuous format is not supported. I have processed my dataset into the same form as MVtec. What is the reason?

AlgirdasKartavicius commented 3 years ago

Did you solve your problem?

renqiqi commented 3 years ago

Did you solve your problem?

modify the code src/main.py:line 155

flatten_gt_mask_list = np.concatenate(gt_mask_list).ravel() flatten_gt_mask_list = flatten_gt_mask_list.astype('uint8')
flatten_score_map_list = np.concatenate(score_map_list).ravel()

AlgirdasKartavicius commented 3 years ago

Thank you.