Richard W. Vorder Bruegge

Distinguishing identical twins by face recognition

The paper measures the ability of face recognition algorithms to distinguish between identical twin siblings. The experimental dataset consists of images taken of 126 pairs of identical twins (252 people) collected on the same day and 24 pairs of identical twins (48 people) with images collected one year apart. In terms of both the number of paris of twins and lapsed time between acquisitions, this is the most extensive investigation of face recognition performance on twins to date. Recognition experiments are conducted using three of the top submissions to the Multiple Biometric Evaluation (MBE) 2010 Still Face Track [1]. Performance results are reported for both same day and cross year matching. Performance results are broken out by lighting conditions (studio and outside); expression (neutral and smiling); gender and age. Confidence intervals were generated by a bootstrap method. This is the most detailed covariate analysis of face recognition of twins to date.

Key Parameters of Face Shape Variation in 3D in a Large Sample

Abstract:  Improvement of methods for evidential facial comparison for the Courts relies on the collection of large databases of facial images that permit the analysis of face shape variation and the development of statistical tools. In this paper, we present a short description and key findings of an anthropometric study of face shape variation in three‐dimensions. We used Statistical Shape Analysis to investigate a large database sample (n = 1968), classified by age and gender. We found that size, shape of the bilateral features and midline contributed successively to overall variation. Face size is associated with age. Sexual dimorphism is evident in size and shape, and shows patterns that affect male and female subjects differently. We anticipate this approach will lend itself to the development of methods for analysis of variation within subject groups and the establishment of the relative uniqueness or abundance of facial measurements within them.

Analysis of Facial Marks to Distinguish Between Identical Twins

Identical twin face recognition is a challenging task due to the existence of a high degree of correlation in overall facial appearance. Commercial face recognition systems exhibit poor performance in differentiating between identical twins under practical conditions. In this paper, we study the usability of facial marks as biometric signatures to distinguish between identical twins. We propose a multiscale automatic facial mark detector based on a gradient-based operator known as the fast radial symmetry transform. The transform detects bright or dark regions with high radial symmetry at different scales. Next, the detections are tracked across scales to determine the prominence of facial marks. Extensive experiments are performed both on manually annotated and on automatically detected facial marks to evaluate the usefulness of facial marks as biometric signatures. Experiment results are based on identical twin images acquired at the 2009 Twins Days Festival in Twinsburg, Ohio. The results of our analysis signify the usefulness of the distribution of facial marks as a biometric signature. In addition, our results indicate the existence of some degree of correlation between geometric distribution of facial marks across identical twins.

Double Trouble: Differentiating Identical Twins by Face Recognition

Facial recognition algorithms should be able to operate even when similar-looking individuals are encountered, or even in the extreme case of identical twins. An experimental data set comprised of 17486 images from 126 pairs of identical twins (252 subjects) collected on the same day and 6864 images from 120 pairs of identical twins (240 subjects) with images taken a year later was used to measure the performance on seven different face recognition algorithms. Performance is reported for variations in illumination, expression, gender, and age for both the same day and cross-year image sets. Regardless of the conditions of image acquisition, distinguishing identical twins are significantly harder than distinguishing subjects who are not identical twins for all algorithms.

Facial marks as biometric signatures to distinguish between identical twins

There exists a high degree of similarity in facial appearance between identical twins that makes it difficult for even the state of the art face matching systems to distinguish between them. Given the consistent increase in the number of twin births in recent decades, there is a need to develop alternate approaches to characterize facial appearance that can address this challenging task that has eluded even humans. In this paper, we investigate the usefulness of facial marks as biometric signatures with focus on the task of distinguishing between identical twins. We define and characterize a set of facial marks that are manually annotated by multiple observers. The geometric distribution of annotated facial marks along with their respective categories is used to characterize twin face images. The analysis is conducted on 295 twin face images acquired at the Twins Days Festival at Twinsburg, Ohio, in 2009. The results of our analysis signify the usefulness of distribution of facial marks as a biometric signature. In addition, contrary to prior research, our results indicate the existence of some degree of correlation between positions of facial marks belonging to identical twins.

An Exploration of Sample Representativeness in Anthropometric Facial Comparison

Abstract:  Faces are assumed to be unique, but their use in court has remained problematic as no method of comparison with known error rates has been accepted by the scientific community. Rather than relying on the assumed uniqueness of facial features, previous research has been directed at estimations of face shape frequency. Here, the influence of age, sex, and ancestry on variation was investigated. Statistical shape analysis was used to examine the necessity for sub‐divisions in forensic comparisons, using a large sample of facial images on which 30 anthropometric landmark points had been placed in 3D. Results showed a clear pattern of separation of the sexes in all age groups, and in different age groups in men. It was concluded that sub‐division of databases by sex will be necessary in forensic comparisons. Sub‐division by age may be necessary in men (although not necessarily in women), and may be necessary by ancestry.

Facial recognition of identical twins

Biometric identification systems must be able to distinguish between individuals even in situations where the bio metric signature may be similar, such as in the case of identical twins. This paper presents experiments done in facial recognition using data from a set of images of twins. This work establishes the current state of facial recognition in regards to twins and the accuracy of current state-of-the art programs in distinguishing between identical twins using three commercial face matchers, Cognitec 8.3.2.0, VeriLook 4.0, and PittPatt 4.2.1 and a baseline matcher employing Local Region PCA. Overall, it was observed that Cognitec had the best performance. All matchers, how ever, saw degradation in performance compared to an experiment where the ability to distinguish unrelated persons was assessed. In particular, lighting and expression seemed to have affected performance the most.

Photographic Identification of Denim Trousers from Bank Surveillance Film

The ability to identify denim trousers (e.g., “blue jeans”) from bank surveillance film through side-by-side comparison is documented herein. Individual identifying characteristics that are generated in the manufacturing process and during normal wear-and-tear may be recognized on denim trousers and in photographs depicting those trousers. These characteristics may include folds, creases, and puckering which manifest themselves as high and low areas (“ridges and valleys”) along and adjacent to the seams and hems. As the jeans are worn and washed over time, the visibility of these ridges and valleys is amplified through abrasion of the ridges, resulting in a loss of dark dye and marked tonal brightening. Given sufficient abrasion, even small ridges may exhibit sufficient contrast against the dye-rich background to be recorded on bank surveillance films, permitting a comparison with trousers recovered from suspects. The application of this technique in a case involving a series of bombings and bank robberies in Spokane. Washington, is discussed.

Human Identification Using Automatic and Semi-Automatically Detected Facial Marks

Continuing advancements in the field of digital cameras and surveillance imaging devices have led law enforcement and intelligence agencies to use analysis of images and videos for the investigation and prosecution of crime. When determining identity from photographic evidence, forensic analysts perform comparison of visible facial features manually, which is inefficient. In this study, we will address research efforts to use facial marks as biometric signatures to distinguish between individuals. We propose two systems to assist forensic analysts during photographic comparison: an improved multiscale facial mark system in which facial marks are detected automatically, and a semi‐automatic facial mark system that integrates human knowledge within the improved multiscale facial mark system. Experiment results employ a high‐resolution time‐elapsed dataset acquired at the University of Notre Dame between 2009 and 2011. The results indicate that the geometric distributions of facial mark patterns can be used to distinguish between individuals.

Noise reduction of video imagery through simple averaging

Examiners in the Special Photographic Unit of the Federal Bureau of Investigation Laboratory Division conduct examinations of questioned photographic evidence of all types, including surveillance imagery recorded on film and video tape. A primary type of examination includes side-by- side comparisons, in which unknown objects or people depicted in the questioned images are compared with known objects recovered from suspects or with photographs of suspects themselves. Most imagery received in the SPU for such comparisons originate from time-lapse video or film systems. In such circumstances, the delay between sequential images is so great that standard image summing and/or averaging techniques are useless as a means of improving image detail in questioned subjects or objects without also resorting to processing-intensive pattern reconstruction algorithms. Occasionally, however, the receipt of real-time video imagery will include a questioned object at rest. In such cases, it is possible to use relatively simple image averaging techniques as a means of reducing transient noise in the images, without further compromising the already-poor resolution inherent in most video surveillance images. This paper presents an example of one such case in which multiple images were averaged to reduce the transient noise to a sufficient degree to permit the positive identification of a vehicle based upon the presence of scrape marks and dents on the side of the vehicle.Examiners in the Special Photographic Unit (SPU) of the Federal Bureau of Investigation (FBI) Laboratory Division conduct examinations of questioned photographic evidence of all types, including surveillance imagery recorded on film and video tape. A primary type of examination includes side-by-side comparisons, in which unknown objects or people depicted in the questioned images are compared with known objects recovered from suspects or with photographs of suspects themselves. Most imagery received in the SPU for such comparisons originate from timelapse video or film systems. In such circumstances, the delay between sequential images is so great that standard image summing and/or averaging techniques are useless as a means of improving image detail in questioned subjects or objects without also resorting to processing-intensive pattern reconstruction algorithms. Occasionally, however, the receipt of real-time video imagery will include a questioned object at rest. In such cases, it is possible to use relatively simple image averaging techniques as a means of reducing transient noise in the images, without further compromising the already-poor resolution inherent in most video surveillance images. This paper presents an example of one such case in which multiple images were averaged to reduce the transient noise to a sufficient degree to permit the positive identification of a vehicle based upon the presence of scrape marks and dents on the side of the vehicle.