Can holistic representations be used for face biometric quality assessment

Biometric quality: a review of fingerprint, iris, and face

Abstract: Biometric systems encounter variability in data that influence capture, treatment, and u-sage of a biometric sample. It is imperative to first analyze the data and incorporate this understanding within the recognition system, making assessment of biometric quality an important aspect of biometrics. Though several interpretations and definitions of quality exist, sometimes of a conflicting nature, a holistic definition of quality is indistinct. This paper presents a survey of different concepts and interpretations of biometric quality so that a clear picture of the current state and future directions can be presented. Several factors that cause different types of degradations of biometric samples, including image features that attribute to the effects of these degradations, are discussed. Evaluation schemes are presented to test the performance of quality metrics for various applications. A survey of the features, strengths, and limitations of existing quality assessment techniques in fingerprint, iris, and face biometric are also presented. Finally, a representative set of quality metrics from these three modalities are evaluated on a multimodal database consisting of 2D images, to understand their behavior with respect to match scores obtained from the state-of-the-art recognition systems. The analysis of the characteristic function of quality and match scores shows that a careful selection of complimentary set of quality metrics can provide more benefit to various applications of biometric quality.

Learning Face Image Quality From Human Assessments

Abstract: Face image quality can be defined as a measure of the utility of a face image to automatic face recognition. In this paper, we propose (and compare) two methods for learning face image quality based on target face quality values from: 1) human assessments of face image quality (matcher-independent) and 2) quality values computed from similarity scores (matcher-dependent). A support vector regression model trained on face features extracted using a deep convolutional neural network (ConvNet) is used to predict the quality of a face image. The proposed methods are evaluated on two unconstrained face image databases, Labeled Faces in the Wild and IARPA Janus Benchmark-A (IJB-A), which both contain facial variations encompassing a multitude of quality factors. Evaluation of the proposed automatic face image quality measures shows we are able to reduce the false non-match rate at 1% false match rate by at least 13% for two face matchers (a commercial off-the-shelf matcher and a ConvNet matcher) by using the proposed face quality to select subsets of face images and video frames for matching templates (i.e., multiple faces per subject) in the IJB-A protocol. To the best of our knowledge, this is the first work to utilize human assessments of face image quality in designing a predictor of unconstrained face quality that is shown to be effective in cross-database evaluation.

Standardization of Face Image Sample Quality

Abstract: Performance of biometric systems is dependent on quality of acquired biometric samples. Poor sample quality is a main reason for matching errors in biometric systems and may be the main weakness of some implementations. This paper proposes an approach for standardization of facial image quality, and develops facial symmetry based methods for the assessment of it by measuring facial asymmetries caused by non-frontal lighting and improper facial pose. Experimental results are provided to illustrate the concepts, definitions and effectiveness.

Face biometric quality assessment via light CNN

Abstract: In this paper, we proposed a novel biometric quality assessment (BQA) method for face images and explored its applications in face recognition. Here, we considered five categories of common homogeneous distortion in video suvillance applications, i.e. low-resolution, blurring, additive Gaussian white noise, salt and pepper noise, and Poisson noise. In the BQA model, we first learnt a classifier to simultaneously predict the categories and degrees of the degradation in a face image. Because the quality labels are often ambiguous and inaccurate, we used a light convolutional neural network with the Max-Feature-Map units to make the BQA model robust to noisy labels. Afterwards, we calculated the biometric quality score by pooling such predictions based on the recognition confidence of each degradation class. Finally, we proposed one promising strategy for developing reliable face recognition systems based on this BQA method. Thorough experiments have been conducted on the CASIA, FLW, and YouTube databases. The results demonstrate the effectiveness of the proposed BQA method.

QFuse: Online learning framework for adaptive biometric system

Abstract: Biometrics, the science of verifying the identity of individuals, is increasingly being used in several applications such as assisting law enforcement agencies to control crime and fraud. Existing techniques are unable to provide significant levels of accuracy in uncontrolled noisy environments. Further, scalability is another challenge due to variations in data distribution with changing conditions. This paper presents an adaptive context switching algorithm coupled with online learning to address both these challenges. The proposed framework, termed as QFuse, uses the quality of input images to dynamically select the best biometric matcher or fusion algorithm to verify the identity of an individual. The proposed algorithm continuously updates the selection process using online learning to address the scalability and accommodate the variations in data distribution. The results on the WVU multimodal database and a large real world multimodal database obtained from a law enforcement agency show the efficacy of the proposed framework.

Histograms of oriented gradients for human detection

Abstract: We study the question of feature sets for robust visual object recognition; adopting linear SVM based human detection as a test case. After reviewing existing edge and gradient based descriptors, we show experimentally that grids of histograms of oriented gradient (HOG) descriptors significantly outperform existing feature sets for human detection. We study the influence of each stage of the computation on performance, concluding that fine-scale gradients, fine orientation binning, relatively coarse spatial binning, and high-quality local contrast normalization in overlapping descriptor blocks are all important for good results. The new approach gives near-perfect separation on the original MIT pedestrian database, so we introduce a more challenging dataset containing over 1800 annotated human images with a large range of pose variations and backgrounds.

Modeling the Shape of the Scene: A Holistic Representation of the Spatial Envelope

Abstract: In this paper, we propose a computational model of the recognition of real world scenes that bypasses the segmentation and the processing of individual objects or regions. The procedure is based on a very low dimensional representation of the scene, that we term the Spatial Envelope. We propose a set of perceptual dimensions (naturalness, openness, roughness, expansion, ruggedness) that represent the dominant spatial structure of a scene. Then, we show that these dimensions may be reliably estimated using spectral and coarsely localized information. The model generates a multidimensional space in which scenes sharing membership in semantic categories (e.g., streets, highways, coasts) are projected closed together. The performance of the spatial envelope model shows that specific information about object shape or identity is not a requirement for scene categorization and that modeling a holistic representation of the scene informs about its probable semantic category.

Modern image quality assessment

Abstract: This book is about objective image quality assessmentwhere the aim is to provide computational models that can automatically predict perceptual image quality. The early years of the 21st century have witnessed a tremendous growth in the use of digital images as a means for representing and communicating information. A considerable percentage of this literature is devoted to methods for improving the appearance of images, or for maintaining the appearance of images that are processed. Nevertheless, the quality of digital images, processed or otherwise, is rarely perfect. Images are subject to distortions during acquisition, compression, transmission, processing, and reproduction. To maintain, control, and enhance the quality of images, it is important for image acquisition, management, communication, and processing systems to be able to identify and quantify image quality degradations. The goals of this book are as follows; a) to introduce the fundamentals of image quality assessment, and to explain the relevant engineering problems, b) to give a broad treatment of the current state-of-the-art in image quality assessment, by describing leading algorithms that address these engineering problems, and c) to provide new directions for future research, by introducing recent models and paradigms that significantly differ from those used in the past. The book is written to be accessible to university students curious about the state-of-the-art of image quality assessment, expert industrial R&D engineers seeking to implement image/video quality assessment systems for specific applications, and academic theorists interested in developing new algorithms for image quality assessment or using existing algorithms to design or optimize other image processing applications.

The CAS-PEAL Large-Scale Chinese Face Database and Baseline Evaluations

Abstract: In this paper, we describe the acquisition and contents of a large-scale Chinese face database: the CAS-PEAL face database. The goals of creating the CAS-PEAL face database include the following: 1) providing the worldwide researchers of face recognition with different sources of variations, particularly pose, expression, accessories, and lighting (PEAL), and exhaustive ground-truth information in one uniform database; 2) advancing the state-of-the-art face recognition technologies aiming at practical applications by using off-the-shelf imaging equipment and by designing normal face variations in the database; and 3) providing a large-scale face database of Mongolian. Currently, the CAS-PEAL face database contains 99 594 images of 1040 individuals (595 males and 445 females). A total of nine cameras are mounted horizontally on an arc arm to simultaneously capture images across different poses. Each subject is asked to look straight ahead, up, and down to obtain 27 images in three shots. Five facial expressions, six accessories, and 15 lighting changes are also included in the database. A selected subset of the database (CAS-PEAL-R1, containing 30 863 images of the 1040 subjects) is available to other researchers now. We discuss the evaluation protocol based on the CAS-PEAL-R1 database and present the performance of four algorithms as a baseline to do the following: 1) elementarily assess the difficulty of the database for face recognition algorithms; 2) preference evaluation results for researchers using the database; and 3) identify the strengths and weaknesses of the commonly used algorithms.

SCface --- surveillance cameras face database

Abstract: In this paper we describe a database of static images of human faces. Images were taken in uncontrolled indoor environment using five video surveillance cameras of various qualities. Database contains 4,160 static images (in visible and infrared spectrum) of 130 subjects. Images from different quality cameras should mimic real-world conditions and enable robust face recognition algorithms testing, emphasizing different law enforcement and surveillance use case scenarios. In addition to database description, this paper also elaborates on possible uses of the database and proposes a testing protocol. A baseline Principal Component Analysis (PCA) face recognition algorithm was tested following the proposed protocol. Other researchers can use these test results as a control algorithm performance score when testing their own algorithms on this dataset. Database is available to research community through the procedure described at www.scface.org .