0
$\begingroup$

If I have an image dataset that consists of "normal", anomalous and ground truth image data, how do I make use of the ground truth data?

To my understanding if I train an unsupervised anomaly detection approach on "normal" images, it can then later predict whether an input image is "normal" or anomalous.

How can I make use of the ground truth data in this context? And what purpose has the ground truth data in anomaly detection in general?

Thanks in advance for any answers! I am relatively new to the field of anomaly detection.

$\endgroup$
0
$\begingroup$

You could have a numeric ground truth such as how similar each image is to the normal subset. Another option may be a ground truth mask which highlights which part of the image is anomalous. If the ground truth is just another image it isn't clear what purpose it would serve.

| cite | improve this answer | |
$\endgroup$
  • $\begingroup$ Yes, the ground truth data tells me which parts of the anomalous images are anomalous. I just do not quite understand how I can use these to contribute to my anomaly detection approach. $\endgroup$ – mar_ey Oct 18 at 10:57

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.