Match Score

[atharvjairath]

Hey! First of all great work, I was wondering if we can get a match score of 2 images from this model?

[Parskatt]

Hi!

Sure, you could for example take the average of the dense confidence. This gives you an estimate of how much the images overlap.

[atharvjairath]

flow, confidence = dkm_model.match(im1, im2)
print(confidence.mean())

you mean something like above? if not please feel free to suggest. Thanks :)

[Parskatt]

@atharvjairath Yeah, that seems like a good first approach. You may want to adjust the confidence to suit your needs. For example thresholding or attenuation might be more suitable, but the raw confidence should also be quite ok.

[atharvjairath]

@Parskatt Cool thank you, will explore further.