Benjamin_Loison opened issue Benjamin_Loison/scipy#2 2024-03-23 17:00:39 +01:00
Missing import
Ease PIL image pixels loading and accessing
Benjamin_Loison opened issue Benjamin_Loison/numpy#1 2024-03-22 17:16:45 +01:00
Simplify np.max([np.max(prnus) + np.max(images)])
Benjamin_Loison created repository Benjamin_Loison/numpy 2024-03-22 17:13:36 +01:00
Benjamin_Loison opened issue Benjamin_Loison/vpv#2 2024-03-22 15:22:15 +01:00
View an image, possibly a numpy array or PIL image, from Python?
Benjamin_Loison opened issue Benjamin_Loison/vpv#1 2024-03-22 15:14:46 +01:00
Propose a snap alternative (flatpak)
Benjamin_Loison created repository Benjamin_Loison/vpv 2024-03-22 15:12:32 +01:00
72e37a252a Add algorithms/distance/rmsdiff.py
cf22ff2694 Add comments
d818714344 CAI over images mean does not seem better
464e43861b Prefer *mean* over *average*
Benjamin_Loison commented on issue Benjamin_Loison/Robust_image_source_identifica…#9 2024-03-22 12:53:48 +01:00
Context-Adaptive Interpolator (CAI)

Should read back theory to make sure I correctly follow it.

Benjamin_Loison commented on issue Benjamin_Loison/Robust_image_source_identifica…#9 2024-03-22 12:51:50 +01:00
Context-Adaptive Interpolator (CAI)
RMS diff with average image = 146.59313795571444
RMS diff with average CAI images = 145.5368224929786
Actual phone index 0, guessed phone index 7 by averages images and 9 by average CAI…
41926663b7 Correct a typo but PRNU classifier still looking random
36f5b69f5e First PRNU classifier acting randomly it seems
Benjamin_Loison commented on issue Benjamin_Loison/Robust_image_source_identifica…#9 2024-03-22 12:45:51 +01:00
Context-Adaptive Interpolator (CAI)

While being nearest to actual PRNU thanks to CAI, it does not seem to help:

RMS diff with average image = 100.26988483431852
RMS diff with average CAI images = 79.12339215595037
RMS diff…
Theoretical correct Gaussian noise image?
a9adf2d53d Compare RMS difference between average images and the average CAI images