Andreas Uhl

Eduardo Ribeiro, Andreas Uhl, Fernando Alonso-Fernandez

The use of low-resolution images adopting more relaxed acquisition conditions such as mobile phones and surveillance videos is becoming increasingly common in iris recognition nowadays. Concurrently, a great variety of single image super-resolution techniques are emerging, especially with the use of convolutional neural networks (CNNs). The main objective of these methods is to try to recover finer texture details generating more photo-realistic images based on the optimisation of an objective function depending basically on the CNN architecture and training approach. In this work, the authors explore single image super-resolution using CNNs for iris recognition. For this, they test different CNN architectures and use different training databases, validating their approach on a database of 1.872 near infrared iris images and on a mobile phone image database. They also use quality assessment, visual results and recognition experiments to verify if the photo-realism provided by the CNNs which have already proven to be effective for natural images can reflect in a better recognition rate for iris recognition. The results show that using deeper architectures trained with texture databases that provide a balance between edge preservation and the smoothness of the method can lead to good results in the iris recognition process.

Heinz Hofbauer, Fernando Alonso-Fernandez, Josef Bigun et al.

In this study the authors will look at the detection and segmentation of the iris and its influence on the overall performance of the iris-biometric tool chain. The authors will examine whether the segmentation accuracy, based on conformance with a ground truth, can serve as a predictor for the overall performance of the iris-biometric tool chain. That is: If the segmentation accuracy is improved will this always improve the overall performance? Furthermore, the authors will systematically evaluate the influence of segmentation parameters, pupillary and limbic boundary and normalisation centre (based on Daugman's rubbersheet model), on the rest of the iris-biometric tool chain. The authors will investigate if accurately finding these parameters is important and how consistency, that is, extracting the same exact region of the iris during segmenting, influences the overall performance.

Eduardo Ribeiro, Andreas Uhl, Fernando Alonso-Fernandez et al.

In this work we test the ability of deep learning methods to provide an end-to-end mapping between low and high resolution images applying it to the iris recognition problem. Here, we propose the use of two deep learning single-image super-resolution approaches: Stacked Auto-Encoders (SAE) and Convolutional Neural Networks (CNN) with the most possible lightweight structure to achieve fast speed, preserve local information and reduce artifacts at the same time. We validate the methods with a database of 1.872 near-infrared iris images with quality assessment and recognition experiments showing the superiority of deep learning approaches over the compared algorithms.

Christian Rathgeb, Jascha Kolberg, Andreas Uhl et al.

Today, deep learning represents the most popular and successful form of machine learning. Deep learning has revolutionised the field of pattern recognition, including biometric recognition. Biometric systems utilising deep learning have been shown to achieve auspicious recognition accuracy, surpassing human performance. Apart from said breakthrough advances in terms of biometric performance, the use of deep learning was reported to impact different covariates of biometrics such as algorithmic fairness, vulnerability to attacks, or template protection. Technologies of biometric template protection are designed to enable a secure and privacy-preserving deployment of biometrics. In the recent past, deep learning techniques have been frequently applied in biometric template protection systems for various purposes. This work provides an overview of how advances in deep learning take influence on the field of biometric template protection. The interrelation between improved biometric performance rates and security in biometric template protection is elaborated. Further, the use of deep learning for obtaining feature representations that are suitable for biometric template protection is discussed. Novel methods that apply deep learning to achieve various goals of biometric template protection are surveyed along with deep learning-based attacks.

Johannes Schuiki, Michael Linortner, Georg Wimmer et al.

The last decade has brought forward many great contributions regarding presentation attack detection for the domain of finger and hand vein biometrics. Among those contributions, one is able to find a variety of different attack databases that are either private or made publicly available to the research community. However, it is not always shown whether the used attack samples hold the capability to actually deceive a realistic vein recognition system. Inspired by previous works, this study provides a systematic threat evaluation including three publicly available finger vein attack databases and one private dorsal hand vein database. To do so, 14 distinct vein recognition schemes are confronted with attack samples and the percentage of wrongly accepted attack samples is then reported as the Impostor Attack Presentation Match Rate. As a second step, comparison scores from different recognition schemes are combined using score level fusion with the goal of performing presentation attack detection.

Simon Kirchgasser, Christof Kauba, Georg Wimmer et al.

Natural Scene Statistics commonly used in non-reference image quality measures and a deep learning based quality assessment approach are proposed as biometric quality indicators for vasculature images. While NIQE and BRISQUE if trained on common images with usual distortions do not work well for assessing vasculature pattern samples' quality, their variants being trained on high and low quality vasculature sample data behave as expected from a biometric quality estimator in most cases (deviations from the overall trend occur for certain datasets or feature extraction methods). The proposed deep learning based quality metric is capable of assigning the correct quality class to the vaculature pattern samples in most cases, independent of finger or hand vein patterns being assessed. The experiments were conducted on a total of 13 publicly available finger and hand vein datasets and involve three distinct template representations (two of them especially designed for vascular biometrics). The proposed (trained) quality measures are compared to a several classical quality metrics, with their achieved results underlining their promising behaviour.

Eduardo Ribeiro, Andreas Uhl, Fernando Alonso-Fernandez

Several recent works have addressed the ability of deep learning to disclose rich, hierarchical and discriminative models for the most diverse purposes. Specifically in the super-resolution field, Convolutional Neural Networks (CNNs) using different deep learning approaches attempt to recover realistic texture and fine grained details from low resolution images. In this work we explore the viability of these approaches for iris Super-Resolution (SR) in an iris recognition environment. For this, we test different architectures with and without a so called image re-projection to reduce artifacts applying it to different iris databases to verify the viability of the different CNNs for iris super-resolution. Results show that CNNs and image re-projection can improve the results specially for the accuracy of recognition systems using a complete different training database performing the transfer learning successfully.

Robert Jöchl, Andreas Uhl

In the context of temporal image forensics, it is not evident that a neural network, trained on images from different time-slots (classes), exploits solely image age related features. Usually, images taken in close temporal proximity (e.g., belonging to the same age class) share some common content properties. Such content bias can be exploited by a neural network. In this work, a novel approach is proposed that evaluates the influence of image content. This approach is verified using synthetic images (where content bias can be ruled out) with an age signal embedded. Based on the proposed approach, it is shown that a deep learning approach proposed in the context of age classification is most likely highly dependent on the image content. As a possible countermeasure, two different models from the field of image steganalysis, along with three different preprocessing techniques to increase the signal-to-noise ratio (age signal to image content), are evaluated using the proposed method.

Robert Jöchl, Andreas Uhl

The goal of temporal image forensic is to approximate the age of a digital image relative to images from the same device. Usually, this is based on traces left during the image acquisition pipeline. For example, several methods exist that exploit the presence of in-field sensor defects for this purpose. In addition to these 'classical' methods, there is also an approach in which a Convolutional Neural Network (CNN) is trained to approximate the image age. One advantage of a CNN is that it independently learns the age features used. This would make it possible to exploit other (different) age traces in addition to the known ones (i.e., in-field sensor defects). In a previous work, we have shown that the presence of strong in-field sensor defects is irrelevant for a CNN to predict the age class. Based on this observation, the question arises how device (in)dependent the learned features are. In this work, we empirically asses this by training a network on images from a single device and then apply the trained model to images from different devices. This evaluation is performed on 14 different devices, including 10 devices from the publicly available 'Northumbria Temporal Image Forensics' database. These 10 different devices are based on five different device pairs (i.e., with the identical camera model).

Robert Jöchl, Andreas Uhl

In this work, a novel multi-face tracking method named FaceQSORT is proposed. To mitigate multi-face tracking challenges (e.g., partially occluded or lateral faces), FaceQSORT combines biometric and visual appearance features (extracted from the same image (face) patch) for association. The Q in FaceQSORT refers to the scenario for which FaceQSORT is desinged, i.e. tracking people's faces as they move towards a gate in a Queue. This scenario is also reflected in the new dataset `Paris Lodron University Salzburg Faces in a Queue', which is made publicly available as part of this work. The dataset consists of a total of seven fully annotated and challenging sequences (12730 frames) and is utilized together with two other publicly available datasets for the experimental evaluation. It is shown that FaceQSORT outperforms state-of-the-art trackers in the considered scenario. To provide a deeper insight into FaceQSORT, comprehensive experiments are conducted evaluating the parameter selection, a different similarity metric and the utilized face recognition model (used to extract biometric features).

Robert Jöchl, Andreas Uhl

Temporal image forensics is the science of estimating the age of a digital image. Usually, time-dependent traces (age traces) introduced by the image acquisition pipeline are exploited for this purpose. In this review, a comprehensive overview of the field of temporal image forensics based on time-dependent traces from the image acquisition pipeline is given. This includes a detailed insight into the properties of known age traces (i.e., in-field sensor defects and sensor dust) and temporal image forensics techniques. Another key aspect of this work is to highlight the problem of content bias and to illustrate how important eXplainable Artificial Intelligence methods are to verify the reliability of temporal image forensics techniques. Apart from reviewing material presented in previous works, in this review: (i) a new (probably more realistic) forensic setting is proposed; (ii) the main properties (growth rate and spatial distribution) of in-field sensor defects are verified; (iii) it is shown that a method proposed to utilize in-field sensor defects for image age approximation actually exploits other traces (most likely content bias); (iv) the features learned by a neural network dating palmprint images are further investigated; (v) it is shown how easily a neural network can be distracted from learning age traces. For this purpose, previous work is analyzed, re-implemented if required and experiments are conducted.

Babak Maser, Andreas Uhl

We study the finger vein (FV) sensor model identification task using a deep learning approach. So far, for this biometric modality, only correlation-based PRNU and texture descriptor-based methods have been applied. We employ five prominent CNN architectures covering a wide range of CNN family models, including VGG16, ResNet, and the Xception model. In addition, a novel architecture termed FV2021 is proposed in this work, which excels by its compactness and a low number of parameters to be trained. Original samples, as well as the region of interest data from eight publicly accessible FV datasets, are used in experimentation. An excellent sensor identification AUC-ROC score of 1.0 for patches of uncropped samples and 0.9997 for ROI samples have been achieved. The comparison with former methods shows that the CNN-based approach is superior and improved the results.

Babak Maser, Andreas Uhl

Identifying the origin of a sample image in biometric systems can be beneficial for data authentication in case of attacks against the system and for initiating sensor-specific processing pipelines in sensor-heterogeneous environments. Motivated by shortcomings of the photo response non-uniformity (PRNU) based method in the biometric context, we use a texture classification approach to detect the origin of finger vein sample images. Based on eight publicly available finger vein datasets and applying eight classical yet simple texture descriptors and SVM classification, we demonstrate excellent sensor model identification results for raw finger vein samples as well as for the more challenging region of interest data. The observed results establish texture descriptors as effective competitors to PRNU in finger vein sensor model identification.

Georg Wimmer, Rudolf Schraml, Heinz Hofbauer et al.

The proof of origin of logs is becoming increasingly important. In the context of Industry 4.0 and to combat illegal logging there is an increasing motivation to track each individual log. Our previous works in this field focused on log tracking using digital log end images based on methods inspired by fingerprint and iris-recognition. This work presents a convolutional neural network (CNN) based approach which comprises a CNN-based segmentation of the log end combined with a final CNN-based recognition of the segmented log end using the triplet loss function for CNN training. Results show that the proposed two-stage CNN-based approach outperforms traditional approaches.

Christof Kauba, Luca Debiasi, Andreas Uhl

Fake fingerprint representation pose a severe threat for fingerprint based authentication systems. Despite advances in presentation attack detection technologies, which are often integrated directly into the fingerprint scanner devices, many fingerprint scanners are still susceptible to presentation attacks using physical fake fingerprint representation. In this work we evaluate five different commercial-off-the-shelf fingerprint scanners based on different sensing technologies, including optical, optical multispectral, passive capacitive, active capacitive and thermal regarding their susceptibility to presentation attacks using fake fingerprint representations. Several different materials to create the fake representation are tested and evaluated, including wax, cast, latex, silicone, different types of glue, window colours, modelling clay, etc. The quantitative evaluation includes assessing the fingerprint quality of the samples captured from the fake representations as well as comparison experiments where the achieved matching scores of the fake representations against the corresponding real fingerprints indicate the effectiveness of the fake representations. Our results confirmed that all except one of the tested devices are susceptible to at least one type/material of fake fingerprint representations.

Georg Wimmer, Michael Gadermayr, Andreas Vécsei et al.

Novel imaging technologies raise many questions concerning the adaptation of computer-aided decision support systems. Classification models either need to be adapted or even newly trained from scratch to exploit the full potential of enhanced techniques. Both options typically require the acquisition of new labeled training data. In this work we investigate the applicability of image-to-image translation to endoscopic images showing different imaging modalities, namely conventional white-light and narrow-band imaging. In a study on computer-aided celiac disease diagnosis, we explore whether image-to-image translation is capable of effectively performing the translation between the domains. We investigate if models can be trained on virtual (or a mixture of virtual and real) samples to improve overall accuracy in a setting with limited labeled training data. Finally, we also ask whether a translation of testing images to another domain is capable of improving accuracy by exploiting the enhanced imaging characteristics.

Christoph Hofer, Roland Kwitt, Marc Niethammer et al.

Inferring topological and geometrical information from data can offer an alternative perspective on machine learning problems. Methods from topological data analysis, e.g., persistent homology, enable us to obtain such information, typically in the form of summary representations of topological features. However, such topological signatures often come with an unusual structure (e.g., multisets of intervals) that is highly impractical for most machine learning techniques. While many strategies have been proposed to map these topological signatures into machine learning compatible representations, they suffer from being agnostic to the target learning task. In contrast, we propose a technique that enables us to input topological signatures to deep neural networks and learn a task-optimal representation during training. Our approach is realized as a novel input layer with favorable theoretical properties. Classification experiments on 2D object shapes and social network graphs demonstrate the versatility of the approach and, in case of the latter, we even outperform the state-of-the-art by a large margin.

Sebastian Hegenbart, Roland Kwitt, Andreas Uhl

The 39th annual workshop of the Austrian Association for Pattern Recognition (OAGM/AAPR) provides a platform for presentation and discussion of research progress as well as research projects within the OAGM/AAPR community.