In cases of digital enrolment via mobile and online services, identity documents (IDs) verification is critical to efficiently detect forgery and therefore build user trust in the digital world. In this paper, we propose a copy-move public dataset, called FMIDV (forged mobile ID video dataset) containing forged IDs with respect to guilloche patterns. Also, we propose two fraud detection models on guilloche patterns of IDs, which are based on contrastive and adversarial learning. In the sequel, each proposed model manages to read the entire ID and to recognize the guilloche pattern to check its similarity to the pattern of an authentic ID. The objective of the similarity check is to validate its authenticity or its rejection. Experiments are conducted on MIDV and FMIDV datasets to analyze and identify the most proper parameters to achieve higher authentication performance. The code and the dataset are available at https://github.com/malghadi/CheckID.

In order to combat fraud, identity documents and currencies often include security elements such as guilloches, micro prints or holograms. This paper aims to authenticate such documents from videos acquired with a smartphone by analyzing the holograms. The proposed method consists of recognizing all the patterns of the hologram to determine if the document is genuine or not. The Local Binary Patterns (LBP) descriptor is used in this paper to represent the features of a hologram. For a given document, Multi LBP Models are built as a reference model. This model is then compared to the LBP models of the tested hologram to decide if the hologram exist or not in the document and then to determine if the document is genuine or not. Experiments are carried out on holograms of French Passports and Euro banknotes. The results show that the proposed strategy allows to determine if the document is an authentic document or falsified in a good accuracy. The code is available at https://github.com/mnchapel/authentication_f_holograms_with_mixed_patterns_by_direct_lbp_comparison.

Face Presentation Attack Detection (PAD) is an important measure to prevent spoof attacks for face biometric systems. Many works based on Convolution Neural Networks (CNNs) for face PAD formulate the problem as an image-level binary classification task without considering the context. Alternatively, Vision Transformers (ViT) using self-attention to attend the context of an image become the mainstreams in face PAD. Inspired by ViT, we propose a Video-based Transformer for face PAD (ViTransPAD) with short/long-range spatio-temporal attention which can not only focus on local details with short-range attention within a frame but also capture long-range dependencies over frames. Instead of using coarse image patches with single-scale as in ViT, we pro-pose the Multi-scale Multi-Head Self-Attention (MsMHSA) module to accommodate multi-scale patch partitions of Q, K, V feature maps to different heads on a single transformer in a coarse-to-fine manner, which enables to learn a fine-grained representation to perform pixel-level discrimination for face PAD. Due to lack inductive biases of convolutions in pure transformers, we also introduce convolutions to our ViTransPAD to integrate the desirable properties of CNNs. The extensive experiments show the effectiveness of our proposed ViTransPAD with a preferable accuracy-computation balance, which can serve as a new backbone for face PAD.

One of important challenges in the document liveness detection process for identity document verification is quality verification. To tackle this challenge, this paper proposes a reference hashing approach to discriminate between the original template of the identity document image and the scan one, which is called CheckScan. Actually, the discrimination process takes place between two aligned identity document images. The proposed approach is made up of two steps: feature extraction based on Fast Fourier Transform (FFT) and hash construction. Feature extraction step involves partitioning the identity document image into set of non-overlapping blocks, and for each block the FFT magnitude spectrum is calculated. Hence, a specific number from the FFT magnitude peaks is selected as discriminative features. The hash construction step quantizes the selected peaks into binary codes by applying a new quantization approach that is based on the coordinates of the selected peaks. These two steps are combined together in this work to achieve good discriminate (well anti-collision) capability for distinct identity document images. Experiments were conducted in order to analyze and identify the most proper parameters to achieve higher discrimination performance. The experimental results were performed on the Mobile Identity Document Video dataset (MIDV-2020), and the results show that the proposed approach builds binary codes quite discriminative for distinct identity document images.

Abstract Ensuring the medical images authenticity becomes an essential need. This need has to keep of course two factors: (i) high robustness and (ii) low complexity in term of processing. A zero‐watermarking scheme could be a practical solution for this matter. This paper proposes a zero‐watermarking algorithm to assure the authenticity of the transmitted medical images through an e‐healthcare network. The targeted image is partitioned into 8 × 8 non‐overlapping blocks, and the Jacobian matrix model is used to construct a meaningful watermark. In order to reduce the complexity, our model does not encrypt the watermark image. The experiment results show a remarkable efficiency of the proposed model in terms of similarity, error probability, and robustness against a variety of geometric and non‐geometric attacks. Copyright © 2017 John Wiley & Sons, Ltd.

The study aims to preserve integrity of JPEG bitstream transmitted in Wireless Multimedia Sensor Networks (WMSNs) environment, lightweight complex process that is proposed to parse bitstream of encoded medical image and formulated the encryption process for that represented bit of Diff value in DC coefficients rather than appended bits of ACs coefficients. Our choice deals with low down complexity by encrypting limited numbers of DC coefficients versus ACs coefficients per block, or even all image blocks. This model may be utilized in building a secure framework of remote image analysis and archiving center for that transmitted medical images from body sensors and clinician, or even other remote imaging center with high integrity and low complexity. The experiments result shows that the proposed approach gives an interesting and remarkable result to preserve the medical image integrity. The obtained results are discussed in details.