Orientador: João Baptista Tadanobu Yabu-uti / Dissertação (mestrado) - Universidade Estadual de Campinas, Faculdade de Engenharia Eletrica e de Computação / Made available in DSpace on 2018-08-06T16:56:48Z (GMT). No. of bitstreams: 1 LarcoBravo_JulioCesar_M.pdf: 1532173 bytes, checksum: 6a43168475c621b64174f99ac42aadc5 (MD5) Previous issue date: 2006 / Resumo: Uma assinatura manuscrita, é a forma mais utilizada para con?rmar a dentidade de uma pessoa, já a que o estilo de assinar de um indivíduo é uma entidade biométrica que pode ser usada para diferenciar uma pessoa de outra. Neste trabalho, apresenta-se uma metodologia para realizar a autentica¸ao pessoal utilizando o som que se produz no momento de assinar.Quando uma pessoa assina, a fricção entre a ponta rígida de uma caneta e o papel produz um som que pode ser usado para veri?car a identidade de uma pessoa. Esta metodologia, está a baseada no fato de que o som produzido ao a correlacionado com a dinâmica e a postura do assinante. Cada um dos traços que compõe a assinatura corresponde a uma parte do sinal de som. Diferentes indivíduos produzem diferentes traços ou garranchos os quais resultam em diferentes sinais de som. Do sinal de som capturado são correspondem ao som da assinatura e calcula-se a envoltória deste sinal. Cada um dos traços que formam parte da assinatura são caracterizados como variações agudas nos valores da envoltória (picos), as quais ao representadas como vetores binários de características que são enviados para uma etapa de reconhecimento de padrões, a qual decidirá se o som capturado provém de uma assinatura que foi realizada por um usuúario legítimo ou por um impostor. A metodologia apresentada é avaliada utilizando um conjunto de amostras de teste e treinamento pertencentes a dois tipos de usuários: legítimo e impostor habilidoso. O usuário legítimo é o proprietário da assinatura e o impostor habilidoso comhece a forma como o usuário legítimo assina. Como parâmetros de avaliação do desempenho desta metodologia, foram obtidas as taxas de erro FAR (falsa aceitação) e FRR (falsa rejeição) de 8,55% e 8,73%, respectivamente / Abstract: A signature is the most used way to validate a person¿s identity, since the style of every individual signature constitutes a biometric entity, whichcan be used to differentiate one person from another. This research work presents a method to accomplish the personal authentication using the sound produced when a person is signing. During this event, the friction produced by the rigid tip of a pen rubbing the paper, generates a sound that can be used to verify the identity of a person. The reliability of this methodology is based on the fact that the sound emitted during the signature action is closely correlated with the dynamics and posture of the person who signs.Moreover, every line ofthe signature corresponds directlyto one partofthe audio signal generated. Therefore, diferent individuals are able to produce completely different traces or scrawls, which will generate different audio signals. Once the audio signal is digitally captured, the samples that do not belong to the signature are discardedandthe envelope ofthis signal is computed. Everyconstituent trace of the signature are characterizedas sharpvariations of the envelope values (peaks), whichare represented as binary vectors of features. This information, is sent to a pattern recognizing stage which has the responsibility to decide whether the captured sound corresponds to an authentic user or an impostor. The presented methodology is evaluated using a group of test and training samples belongingto two types ofusers: legitimate and skilled impostor.The legitimate user is the proprietor of the signature and the skilled impostor knows the form as the legitimate user signs. As parameters of evaluation of this methodology, were obtained the error rate FAR (false acceptance) and FRR (false rejection) of 8,55% and 8,73%, respectively / Mestrado / Mestre em Engenharia Elétrica

Cryptographic credentials allow user authorizations to be granted and verified. and have such applications as e-Passports, e-Commerce, and electronic cash. This thesis proposes a privacy protecting approach of binding biometrically derived keys to cryptographic credentials to prevent unauthorized lending. Our approach builds on the 2011 work of Adams, offering additional benefits of privacy protection of biometric information, generality on biometric modalities, and performance. Our protocol integrates into Brands’ Digital Credential scheme, and the Anonymous Credentials scheme of Camenisch and Lysyanskaya. We describe a detailed integration with the Digital Credential Scheme and sketch the integration into the Anonymous Credentials scheme. Security proofs for non-transferability, correctness of ownership, and unlinkability are provided for the protocol’s instantiation into Digital Credentials. Our approach uses specialized biometric devices in both the issue and show protocols. These devices are configured with our proposed primitive, the fuzzy ex-tractor indistinguishability adaptor which uses a traditional fuzzy extractor to create and regenerate cryptographic keys from biometric data and IND-CCA2 secure en-cryption protect the generated public data against multiplicity attacks. Pedersen commitments are used to hold the key at issue and show time, and A zero-knowledge proof of knowledge is used to ensure correspondence of key created at issue-time and regenerated at show-time. The above is done in a manner which preserves biometric privacy, as and delivers non-transferability of digital credentials. The biometric itself is not stored or divulged to any of the parties involved in the protocol. Privacy protection in multiple enrollments scenarios is achieved by the fuzzy extractor indistinguishability adapter. The zero knowledge proof of knowledge is used in the showing protocol to prove knowledge of values without divulging them.

This dissertation project takes an interdisciplinary approach towards theorizing how we understand new modes of incarceration and confinement in the digital age. It makes key interventions in the fields of surveillance studies, carceral studies, critical data and technology studies, ethnic and racial studies. I argue that less conventional modes of incarceration and confinement, which are enabled through technologies, the Internet and processes of datafication, conceal the everyday carceral functions that target and exploit racialized people. Chapter 1 examines mobile carceral technologies that are part of Canada’s immigration and detention system. I investigate how notions of increased freedom that are associated with carceral technologies like electronic monitoring and voice reporting do not necessarily coincide with increased autonomy. In Chapter 2, I consider the relationship between mobile phone cameras and the rise of police body-worn cameras. More specifically, I examine how policing and surveillance technologies disproportionately take aim at Black people and communities, making the mere occupation of public and digital space extremely precarious. Lastly, in Chapter 3, I challenge the notion that biometric systems and technologies are race-neutral guarantors of identity, specifically within the polemical space of the modern airport. I argue that the airport’s security and surveillance infrastructure operates according to racialized knowledges, which unofficially validate the profiling of Muslim travelers by both human and non-human operators. / Dissertation / Doctor of Philosophy (PhD) / This dissertation encourages the reader to rethink notions of incarceration from both theoretical and practical perspectives; however, it is not a project about incarceration in the traditional sense. I argue that any notion of incarceration needs to be re-conceptualized in an age that is driven by big data and emergent technologies. While I draw on state and institutional forms of confinement in Canada, all of which have long and established histories of racism and oppression, I contend that notions of incarceration or confinement have bled into everyday life, particularly for racialized and marginalized people and communities. By surveying different surveillance technologies deployed across Canada’s immigration and detention system, the institution of policing and the biometric airport, I suggest that our understanding of the carceral has drastically changed. As issues of race, discrimination and oppression continue to underpin the structures of this newer carceral system and its modes of surveillance and confinement, it is a system that is less visible and physically confining but equally restrictive.

This work addresses the problem of recovering the morphology of blood volume pulse (BVP) information from a video of a person's face. Video-based remote plethysmography methods have shown promising results in estimating vital signs such as heart rate and breathing rate. However, recovering the instantaneous pulse rate signals is still a challenge for the community. This is due to the fact that most of the previous methods concentrate on capturing the temporal average of the cardiovascular signals. In contrast, we present an approach in which BVP signals are extracted with a focus on the recovery of the signal morphology as a generalized form for the computation of physiological metrics. We also place emphasis on allowing natural movements by the subject. Furthermore, our system is capable of extracting individual BVP instances with sufficient signal detail to facilitate candidate re-identification. These improvements have resulted in part from the incorporation of a robust skin-detection module into the overall imaging-based photoplethysmography (iPPG) framework. We present extensive experimental results using the challenging UBFC-Phys dataset and the well-known COHFACE dataset. The source code is available at https://github.com/yogeshd21/CVPM-2023-iPPG-Paper. / Master of Science / In this work we are trying to study and recover human health related metrics and the physiological signals which are at the core for the derivation of such metrics. A well know form of physiological signals is ECG (Electrocardiogram) signals and for our research we work with BVP (Blood Volume Pulse) signals. With this work we are proposing a Deep Learning based model for non-invasive retrieval of human physiological signals from human face videos. Most of the state of the art models as well as researchers try to recover averaged cardiac pulse based metrics like heart rate, breathing rate, etc. without focusing on the details of the recovered physiological signal. Physiological signals like BVP have details like systolic peak, diastolic peak and dicrotic notch, and these signals also have applications in various domains like human mental health study, emotional stimuli study, etc. Hence with this work we focus on retrieval of the morphology of such physiological signals and present a quantitative as well as qualitative results for the same. An efficient attention based deep learning model is presented and scope of reidentification using the retrieved signals is also explored. Along with significant implementations like skin detection model our proposed architecture also shows better performance than state of the art models for two very challenging datasets UBFC-Phys as well as COHFACE. The source code is available at https://github.com/yogeshd21/CVPM-2023-iPPG-Paper.

No / Facial expressions have been studied looking for its diagnostic capabilities in mental health and clues for longevity, gender and other such personality traits. The use of facial expressions, especially the expression of smile, as a biometric has not been looked into great detail. However, research shows that a person can be identified from their behavioural traits including their emotional expressions. In this Chapter, we discuss a novel computational biometric model which can be derived from the smile expression. We discuss how the temporal components of a smile can be utilised to show that similarities in the smile exist for an individual and it can be enabled to create a tool which can be utilised as a biometric.

<p> Biometrics is the science of measuring either physiological (i.e. fingerprint, iris) or behavioural (i.e. gait, signature) characteristics for the purpose of determining or authenticating one's identity. While there has been considerable research conducted with respect to the technical aspects of biometrics, very little attention has been paid to consumer acceptability of this technology. The research presented here is a first step towards filling that void.</p> <p> As such, a series of three studies were undertaken. The first study was a qualitative analysis that identified what avenues of exploration Canadian banks considered to be the most salient with respect to consumer perceptions of biometric authentication technology. This analysis consisted of semi-structured interviews with subject matter experts. The second study was also qualitative and asked consumers from across Canada what they perceived as potential benefits and concerns with biometric authentication technology being used to access their bank accounts. Based upon the results of these two studies, which were further informed by a review of technology adoption literature, a third quantitative study was carried out in which a proposed research model was tested. This model identified both contextual antecedents and innate traits that may influence consumer attitudes towards using biometrics to access their bank accounts via an automated teller machine (ATM). In addition, the aspects of control and voluntariness were manipulated, through the presentation of various scenarios, to examine their effects upon both attitude as well as the direct antecedents of privacy and security concerns and usefulness. The proposed model was assessed using structural equation modeling. In addition, ANOV As and qualitative answers to open ended questions were examined to provide further insight as to what will enhance or impede consumer acceptance of biometric technology.</p> <p> The findings suggest that the contextual factors of privacy and security concerns and usefulness have a bigger impact upon attitude as compared to innate personality traits. In addition, while voluntariness appears to have no effect, control has a significant impact upon attitude as well as privacy and security concerns and usefulness. Based upon these results, implications for theory and practice are discussed, and suggestions for future research are presented. It is hoped that this initial research spurs additional interest in examining consumer acceptability of biometrics in terms of both private and public sector applications.</p> / Thesis / Doctor of Philosophy (PhD)

Postmortem fingerprint collection is a common practice at medical examiners’ offices. Fingerprints are often collected with electronic scanners or ink pads and fingerprint cards. However, obstacles to obtaining clear impressions such as rigor mortis and decomposition can be difficult to overcome using the current methods. There is no clear best method for collecting these compromised fingerprints. The GelSight Mobile™ is a handheld three-dimensional contact imaging system that can measure the topography of any surface regardless of the lighting conditions of the environment. The resolution of the images created is extremely high and can be used to measure single micron features. The goal of this project was to determine if the GelSight Mobile™ was a suitable method for postmortem fingerprint exemplar collection, and to determine if it provides a higher quality fingerprint impression than current postmortem fingerprint collection methods. For this study, three methods – black ink, two-dimensional scanning, and the GelSight Mobile™ – were used on decedents with varying ranges of decomposition to determine the best method for postmortem fingerprint collection. The postmortem interval for the decedents ranged from one day to almost one year, with the latter being exposed to outdoor environments for approximately two weeks prior to discovery and then stored for over a year. Embalmed cadavers were also examined. The results revealed that the GelSight Mobile™ captured fingerprints of higher quality, specifically with higher percentages of prints with level three detail and higher counts of minutiae characteristics than the other methods. However, to be optimized for forensic fingerprint collection, it is recommended that the GelSight Mobile™ be adapted to incorporate a larger gel cartridge and software capabilities to include a mirrored image option and a filter to give images an ink-like appearance.

Auditory Evoked Potentials (AEPs) have recently gained attention as a biometric feature that may improve security and address reliability shortfalls of other commonly-used biometric features. The objective of this thesis is to investigate the accuracy with which subjects can be automatically identified or authenticated with machine learning (ML) techniques using a type of AEP known as the speech-evoked frequency following response (FFR). Accordingly, the results show more accurate discrimination between FFRs from different subjects than what has been reported in past studies. The accuracy improvement is searched either by optimized hyperparameter tuning of the ML model or extracting new features from FFRs and feeding them as inputs to the model. Finally, the accuracy of authenticating subjects using FFRs is investigated using a "sheep vs. wolves" scenario. The results of this work shed more light on the potential of use of speech-evoked FFRs in biometric identification and authentication systems.

No description available.

This dissertation reports the development of a technique for automated face recognition, using color images. One of the more powerful techniques for recognition of faces in monochromatic images has been extended to color by the use of hypercomplex numbers called quaternions. Two software implementations have been written of the new method and the analogous method for use on monochromatic images. Test results show that the new method is superior in accuracy to the analogous monochrome method. Although color images are generally collected, the great majority of published research efforts and of commercially available systems use only the intensity features. This surprising fact provided motivation to the three thesis statements proposed in this dissertation. The first is that the use of color information can increase face recognition accuracy. Face images contain many features, some of which are only easily distinguishable using color while others would seem more robust to illumination variation when color is considered. The second thesis statement is that the currently popular technique of graph-based face analysis and matching of features extracted from application of a family of Gabor filters can be extended to use with color. A particular method of defining a filter appropriate for color images is used; the usual complex Gabor filter is adapted to the domain of quaternions.. Four alternative approaches to the extension of complex Gabor filters to quaternions are defined and discussed; the most promising is selected and used as the basis for subsequent implementation and experimentation. The third thesis statement is that statistical analysis can identify portions of the face image that are highly relevant — i.e., locations that are especially well suited for use in face recognition systems. Conventionally, the Gabor-based graph method extracts features at locations that are equally spaced, or perhaps selected manually on a non-uniform graph. We have defined a relevance image, in which the intensity values are computed from the intensity variance across a number of images from different individuals and the mutual information between the pixel distributions of sets of images from different individuals and the same individual. A complete software implementation of the new face recognition method has been developed. Feature vectors called jets are extracted by application of the novel quaternion Gabor filter, and matched against models of other faces. In order to test the validity of the thesis statements, a parallel software implementation of the conventional monochromatic Gabor graph method has been developed and side-by-side testing has been conducted. Testing results show accuracy increases of 3% to 17% in the new color-based method over the conventional monochromatic method. These testing results demonstrate that color information can indeed provide a significant increase in accuracy, that the extension of Gabor filters to color through the use of quaternions does give a viable feature set, and that the face landmarks chosen via statistical methods do have high relevance for face discrimination. / Ph. D.