Overview Repositories
65
 Projects Packages Public Activity Starred Repositories
Benjamin_Loison commented on issue Benjamin_Loison/Robust_image_source_identifica…#24 2024-04-05 11:33:10 +02:00
Make execution reproducible

WebArchive issue for 4d267bf540/articles/Positional%20…

4d267bf540 Add articles/Positional Learning Quentin Bammey/
Benjamin_Loison commented on issue Benjamin_Loison/Robust_image_source_identifica…#30 2024-04-04 12:21:46 +02:00
Estimate PRNU of devices on actual dataset and evaluate our method 
cut -d ',' -f 66 RAISE_all.csv
Benjamin_Loison commented on issue Benjamin_Loison/Robust_image_source_identifica…#33 2024-04-04 11:59:40 +02:00
Is img_as_float more precise than .convert('F')

Seems to have possibly interesting PRNU pattern:

606821d45e Add show_mean_noise.py
Does not 256 precision per channel is too weak to determine the PRNU even with multiple images?
Benjamin_Loison pushed to master at Benjamin_Loison/Linguee_webscraper 2024-04-03 19:52:11 +02:00
7cffb7e37a Make connection more reliable
0e871cf913 Assume that actual entries are prefered by Linguee
Compare 2 commits »
b354f59ac5 Make parallel production ready
0f394faa61 Multicores prototype
faed1e33e8 #33: Add datasets/raise/extract_noise.py
Compare 3 commits »
Is img_as_float more precise than .convert('F')
4dd52aae90 Use plt.imsave instead of toPilImage
Benjamin_Loison commented on issue Benjamin_Loison/Robust_image_source_identifica…#32 2024-04-03 16:53:38 +02:00
Is imagePrnuEstimateNpArray = imageNpArray - denoise_tv_chambolle correct?

Our current imagePrnuEstimateNpArray = imageNpArray - denoise_tv_chambolle, see:

-----BEGIN PGP MESSAGE-----

hF4DTQa9Wom5MBgSAQdAetV5J8WNto2ugE/PErI0skmigfRJX0AMp00hBDAGrCkw
+/2AfPggU…
Is imagePrnuEstimateNpArray = imageNpArray - denoise_tv_chambolle correct?
Is imagePrnuEstimateNpArray = imageNpArray - denoise_tv_chambolle correct?
d3af12ce3e Try to have clear figure for thesis audition
Benjamin_Loison commented on issue Benjamin_Loison/Robust_image_source_identifica…#31 2024-04-03 15:13:51 +02:00
Given device split into two groups and compare both estimated PRNUs and how they evolved when consider more and more images to estimate them within each group

Should print a 2D figure with the evolving PRNU (with all images for instance) with for each x value a representation of the distribution of PRNU pixel values to see if the PRNU does not converge…

Benjamin_Loison commented on issue Benjamin_Loison/Robust_image_source_identifica…#30 2024-04-03 14:53:55 +02:00
Estimate PRNU of devices on actual dataset and evaluate our method

Maybe pay attention to camera settings potentially affecting the PRNU computation.

Benjamin_Loison commented on issue Benjamin_Loison/Robust_image_source_identifica…#30 2024-04-03 01:02:34 +02:00
Estimate PRNU of devices on actual dataset and evaluate our method

How many images per device is there?

Benjamin_Loison commented on issue Benjamin_Loison/Robust_image_source_identifica…#30 2024-04-03 00:18:10 +02:00
Estimate PRNU of devices on actual dataset and evaluate our method

A 2D table having a colormap and actual accuracy values written in each cell to show the accurracy of our method for all values for both number of images to learn the PRNU and to evalute it.

Benjamin_Loison commented on issue Benjamin_Loison/Robust_image_source_identifica…#31 2024-04-02 23:58:57 +02:00
Given device split into two groups and compare both estimated PRNUs and how they evolved when consider more and more images to estimate them within each group

Figure_1

aafb7ebc92 Plot an exhaustive curve to make sure that results are not just by chance