Spoofing PRNU Patterns of Iris Sensors while Preserving Iris Recognition

Abstract

The principle of Photo Response Non-Uniformity (PRNU) is used to link an image with its source, i.e., the sensor that produced it. In this work, we investigate if it is possible to modify an iris image acquired using one sensor in order to spoof the PRNU noise pattern of a different sensor. In this regard, we develop an image perturbation routine that iteratively modifies blocks of pixels in the original iris image such that its PRNU pattern approaches that of a target sensor. Experiments indicate the efficacy of the proposed perturbation method in spoofing PRNU patterns present in an iris image whilst still retaining its biometric content.

NoiseScope: Detecting Deepfake Images in a Blind Setting

“…To address this, our clustering scheme is designed to be agnostic to the GAN(s) used, and is able to extract available fingerprints, even from multiple models. 4 ○ Method should work for images with any type of high-level content (images of faces, animals, objects, etc.) To address this challenge, we use residual image extraction schemes that can effectively suppress high-level content.…”

“…Each time a cluster is updated, we compute a fingerprint (using cluster members), and two clusters are merged if the PCE correlation between their fingerprints is greater than T mer дe . 4 This is done iteratively at each step to grow clusters. By computing correlations using fingerprints, we reduce the risk of random noise impacting our correlation estimates.…”

“…Fingerprint spoofing attack aims to disguise fake images to be from a specific camera device. This attack is commonly studied in the camera fingerprinting literature [4,44,81]. We use the StyleGAN-Face1 dataset to evaluate this countermeasure.…”

“…NoiseScope can detect fake images with up to 99.68% F1 score. 4 ○ Lastly, we extensively evaluate NoiseScope against a variety of countermeasures by assuming an attacker who is aware of NoiseScope's detection pipeline.…”

Detection of Face Morphing Attacks Based on PRNU Analysis

“…Different approaches are proposed in literature to detect such intentional counter forensic attacks, e.g., the "Triangle Test" [92] and more recently Sameer et al [93] proposed a deep learning based CNN model for the detection of counter forensic images. In biometrics, forging of the image identity has only been investigated for iris sensor data by Banerjee et al, [94] and Uhl and Höller [95], where the detection of such attacks is furthermore evaluated in the latter.…”