when i test，the time is so long

[leolv131]

when i test some images， whatever the number of images is 1 or 100，the time spent is so long. I want to know whether it is nomal

[DeepKnowledge1]

No, it is not normal,

[leolv131]

No, it is not normal, i find the time is spent follow the code， my H*W is 4096，it has to loop 4096 times， is it normal? for i in range(H * W): mean = train_outputs[0][:, i] conv_inv = np.linalg.inv(train_outputs[1][:, :, i]) dist = [mahalanobis(sample[:, i], mean, conv_inv) for sample in embedding_vectors] dist_list.append(dist)

[DeepKnowledge1]

Yes, please have a look on the Inference time issue

https://github.com/xiahaifeng1995/PaDiM-Anomaly-Detection-Localization-master/issues/8#issuecomment-1002495651

[DeepKnowledge1]

Read the whole issue