We are happy that our paper titled “Training Data Improvement for Image Forgery Detection using Comprint” was accepted at the IEEE International Conference on Consumer Electronics (ICCE) 2023, held in conjuction with the Consumer Electronics Show (CES). This work was made in collaboration with the Image Processing Research Group of the University of Napels Federico II, Italy.
Hannes Mareen presented this research on 7 January 2023 at ICCE in Las Vegas, Nevada, USA. The poster can be viewed below (full resolution can be found here).
The poster presented at ICCE 2023. Full resolution can be found here.
Manipulated images are a threat to consumers worldwide, when they are used to spread disinformation. Therefore, Comprint enables forgery detection by utilizing JPEG-compression fingerprints. For more detailed insights on Comprint, check out our previous work (including video presentation).
An image is transformed to a comprint or compression fingerprint, that can be used to create a heatmap that detects and localizes the forgeries. For example, the comprint will differ for a region that was previously compressed with a different JPEG Quality Factor (QF).
This paper evaluates the impact of the training set on Comprint’s performance. Most interestingly, we found that including images compressed with low quality factors during training does not have a significant effect on the accuracy, whereas incorporating recompression boosts the robustness.
As such, consumers can use Comprint on their smartphones to verify the authenticity of images.
The source code of Comprint is available on GitHub.com.
Paper: Training Data Improvement for Image Forgery Detection using Comprint
