Submitted by Marlene Santos (marlene.bc.ufg@gmail.com) on 2016-09-15T17:16:01Z No. of bitstreams: 2 Dissertação - Deborah Pereira Carvalho - 2016.pdf: 1291033 bytes, checksum: 11909a32b04cb17d98b30025fea8cce2 (MD5) license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) / Approved for entry into archive by Luciana Ferreira (lucgeral@gmail.com) on 2016-09-16T12:17:36Z (GMT) No. of bitstreams: 2 Dissertação - Deborah Pereira Carvalho - 2016.pdf: 1291033 bytes, checksum: 11909a32b04cb17d98b30025fea8cce2 (MD5) license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) / Made available in DSpace on 2016-09-16T12:17:36Z (GMT). No. of bitstreams: 2 Dissertação - Deborah Pereira Carvalho - 2016.pdf: 1291033 bytes, checksum: 11909a32b04cb17d98b30025fea8cce2 (MD5) license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) Previous issue date: 2016-03-11 / Research with use of exogenous enzymes in feed for broiler, has shown its advantages in improving the digestibility and nutrient availability for birds, resulting in better performance. The objective of this study was to evaluate the effect of adding "on the top" the enzyme protease 0.05% (Cybenza DP 100) in diets for broilers containing animal meal. Two experiments were conducted divided into 4 treatments each experiment. In both experiments were using the same treatments, the treatments offered were: T1 - corn-based feed and soybean meal, T2 - corn-based feed and soybean meal with addition of protease, T3 - based feed corn and soybean meal with addition of meat, feathers and viscera, T4 - corn-based feed and soybean meal with addition of meat, feathers and viscera, with the addition of protease. The "experiment 1" was carried out with 320 chicks Cobb500 strain of 1 to 21 days. The experimental design was completely randomized, divided into batteries, with 4 treatments and 8 repetitions each containing 10 broiler per cage. The variables analyzed were: digestibility and performance variables. The "Experiment 2" was carried out with 720 chicks Cobb500 strain of 1 to 42 days. The experimental design was completely randomized, divided into box with 4 treatments and 6 repetitions each containing 30 broiler per box. The analyzed variables were: performance and Biometry of the gastrointestinal tract organs and the intestines and bowel. In all treatments the data were subjected to analysis of variance, and means were compared by Tukey test at 5% significance. In the " experiment 1 " it was observed that supplementation of the enzyme protease improved the feed conversion only in the pre- initial stage , but proved advantageous to use also in the initial phase , because even with the reduction in nutrient levels the results did not differ . In the " second experiment " it was observed that the use of animal meal for broilers from 1 to 21 days of age proved advantageous , since treatments with animal meal in their composition showed better weight gain and lower feed intake . The protease enzyme supplementation improved viability. / Pesquisas com a utilização de enzimas exógenas na alimentação de frangos de corte, tem evidenciado as suas vantagens em melhorar a digestibilidade e disponibilidade de nutrientes para os frangos, resultando em melhor desempenho. Objetivou-se com o presente trabalho avaliar o efeito da adição “on the top” da enzima protease 0,05% (Cybenza DP 100) em dietas para frangos de corte contendo farinhas de origem animal. Foram conduzidos dois experimentos divididos em 4 tratamentos cada experimento. Em ambos os experimentos foram utilizando os mesmos tratamentos, os tratamentos oferecidos foram: T1 - ração à base de milho e farelo de soja, T2 - ração à base de milho e farelo de soja, com adição de protease, T3 - ração à base de milho e farelo de soja com inclusão de farinha de carne, penas e vísceras, T4 - ração à base de milho e farelo de soja com inclusão de farinha de carne, penas e vísceras, com adição de protease. O “experimento 1” foi realizado com 320 pintos da linhagem Cobb500 de 1 a 21 dias. O delineamento experimental foi inteiramente casualizado, dividido em baterias, com 4 tratamentos e 8 repetições cada, contendo 10 frangos por gaiola. As variáveis analisadas foram: digestibilidade e variáveis de desempenho. O “experimento 2” foi realizado com 720 pintos da linhagem Cobb500 de 1 a 42 dias. O delineamento experimental adotado foi inteiramente casualizado, dividido em box, com 4 tratamentos e 6 repetições cada, contendo 30 frangos por box. As Variáveis analisadas foram: desempenho e Biometria de órgãos do trato gastrointestinal e do intestino e do intestino. Em todos os tratamentos os dados foram submetidos a análise de variância, e as médias foram comparadas pelo teste de Tukey a 5% de significância. No “experimento 1” foi observado que a suplementação da enzima protease melhorou a conversão alimentar apenas na fase pré-inicial, porém se mostrou vantajoso a sua utilização também na fase inicial, pois mesmo com a redução dos níveis nutricionais os resultados não diferiram entre si. No “experimento 2” foi observado que a utilização da farinha de origem animal para frangos de corte de 1 a 21 dias de idade se mostrou vantajosa, uma vez que os tratamentos que receberam farinhas de origem animal em sua composição apresentaram um melhor ganho de peso e menor consumo de ração. A suplementação da enzima protease melhorou a viabilidade.

Submitted by Cássia Santos (cassia.bcufg@gmail.com) on 2017-03-28T11:45:59Z No. of bitstreams: 2 Dissertação - João Neto Cronemberger Galvão - 2017.pdf: 3677244 bytes, checksum: 131ab2bd6a3da711a718560ca96c89b6 (MD5) license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) / Approved for entry into archive by Luciana Ferreira (lucgeral@gmail.com) on 2017-03-28T13:49:38Z (GMT) No. of bitstreams: 2 Dissertação - João Neto Cronemberger Galvão - 2017.pdf: 3677244 bytes, checksum: 131ab2bd6a3da711a718560ca96c89b6 (MD5) license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) / Made available in DSpace on 2017-03-28T13:49:38Z (GMT). No. of bitstreams: 2 Dissertação - João Neto Cronemberger Galvão - 2017.pdf: 3677244 bytes, checksum: 131ab2bd6a3da711a718560ca96c89b6 (MD5) license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) Previous issue date: 2017-02-17 / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior - CAPES / Over the years, biometrics has become a valuable resource for human identification be- cause it is an affordable, safe and reliable technique. Various security entities throughout the world, public and private, are committed to the use of biometrics in security; both in access control and for identifying suspects. In addition, many crimes can be prevented with tools that allow crosschecking of biometric data. However, biometric systems cur- rently in use are developed and provided by companies as proprietary systems. For this reason, the creation of a free and open tool will bring more security, control and savings for the government. For this reason, the objective of this work is to create a prototype of a free and open tool, developed within the Web Services approach, associated to the development of an API that will aim at a standardization of biometrics services, conside- ring the exchange of data between distinct and independent biometric applications with a low level of coupling and interoperability between systems of different platforms and technologies, so that it is possible to add to the service and use of multiple biometric functionalities such as face recognition, voice, fingerprints, hand geometry, iris, or even a combination, among others, to perform a task of personal identification. / A cada dia que passa, a biometria vem se tornando um recurso valioso para identifica- ção humana, pois trata-se de uma técnica acessível, segura e confiável. Várias entidades de segurança, por todo o mundo, públicas e privadas, estão empenhadas na utilização da biometria na área de segurança, tanto no controle de acesso, quanto para identifica- ção de investigados. Além disso, diversos crimes podem ser evitados com ferramentas que possibilitam o cruzamento de dados biométricos. No entanto, a maioria dos sistemas biométricos em uso atualmente são desenvolvidos e fornecidos por empresas como siste- mas proprietários, desenvolvidos isoladamente, sem o compartilhamento de informações e base de dados, não atendendo, a contento, as necessidades de agentes governamentais no que se refere à escala, controle e economia no uso das aplicações de biometria. Por esta razão, o objetivo deste trabalho será a criação de um protótipo de uma ferramenta livre e aberta, desenvolvida dentro da abordagem de Serviços Web, associado ao desen- volvimento de uma API que visará a padronização dos serviços de biometria, tendo em vista o intercâmbio de dados entre aplicações biométricas distintas e independentes, com um nível reduzido de acoplamento e com interoperabilidade entre sistemas de diferentes plataformas e tecnologias, de modo que seja possível agregar ao serviço o uso de várias funcionalidades biométricas tais como: o reconhecimento de face, a voz, as impressões digitais, a geometria da mão, a íris, ou mesmo a combinação destas, dentre outras, para executar a tarefa de identificação pessoal.

This master’s thesis work was performed at Optimum Biometric Labs, OBL, located in Karlskrona, Sweden. Optimum Biometric Labs perform independent scenario evaluations to companies who develop biometric devices. The company has a product Optimum preConTM which is surveillance and diagnosis tool for biometric systems. This thesis work’s objective was to develop a conceptual model and implement it as an additional layer above the biometric layer with intelligence about the biometric users. The layer is influenced by the general procedure of biometrics in a multimodal behavioural way. It is working in an unsupervised way and performs in an unsupervised manner. While biometric systems are increasingly adopted the technologies have some inherent problems such as false match and false non-match. In practice, a rejected user can not be interpreted as an impostor since the user simply might have problems using his/her biometric feature. The proposed methods in this project are dealing with these problems when analysing biometric usage in runtime. Another fact which may give rise to false rejections is template aging; a phenomenon where the enrolled user’s template is too old compared towards the user’s current biometric feature. A theoretical approach of template aging was known; however since the analysis of template aging detection was correlated with potential system flaws such as device defects or human generated risks such as impostor attacks this task would become difficult to solve in an unsupervised system but when ignoring the definition of template aging, the detection of similar effects was possible. One of the objectives of this project was to detect template aging in a predictive sense; this task failed to be carried out because the absence of basis performing this kind of tasks. The developed program performs abnormality detection at each incoming event from a biometric system. Each verification attempt is assumed to be from a genuine user unless any deviation according to the user's history is found, an abnormality. The possibility of an impostor attack depends on the degree of the abnormality. The application makes relative decisions between fraud possibilities or if genuine user was the source of what caused the deviations. This is presented as an alarm with the degree of impostor possibility. This intelligent layer has increased Optimum preCon´s capacity as a surveillance tool for biometrics. This product is an efficient complement to biometric systems in a steady up-going worldwide market.

Context. Biometric technology is a secure and convenient identification method and it does not need to remember complex passwords, nor smart cards, keys, and the like. Biometrics is the measurable characteristics of individuals based on their behavioral patterns or physiological features that can be used to verify or recognize their identity. Physical characteristics include fingerprints, palm or hand geometry, iris, retina, and facial characteristics. Behavioral characteristics include signature, keystroke and voice pattern. With the combination of biometric technology products and modern computer technology, it is easy to perform monitoring, management, systems integration, automated management, and security applications. Objective. The aim of this research is to explore and clarify the main influencing factors and attitudes concerning biometrics Security technology by analyzing expert opinions. This is done through informal interviews and a web based survey in Blekinge healthcare. Methods. Literature review was the starting phase to map the current state of research in biometric technology implementation. The literature review helped authors to explore and solve different ambiguities in authors’ minds, related to work flow, methods and procedures for different tasks. In our research, we selected five different interviewees from biometric solution supplier companies in Sweden and Denmark, Blekinge healthcare staff, biometric technology technical staff and IT security concerned to express their experiences, opinions and domain knowledge about the implementation of biometrics system in the county of Blekinge. Due to the resource and limited time constraint authors selected survey as a data collection procedure. In survey we planned a questionnaire with the different people related to healthcare. The questionnaire purpose was to collect the quantitative data and our questionnaire was relying on close ended questions. Results. It is analyzed that people have trust on biometrics system and in future they are agreed to face changing in the current system as well as the people who are related with healthcare system are already aware about biometrics and they will easily adopt the implementation of biometric system in healthcare. Conclusion. In concern of user acceptance it is analyzed that people have trust on biometrics system and in future they are agreed to face changing in the current system. In concern of implementation authors analyzed that it is very necessary to conduct a study in order to analyze the requirement of different actors that will participate in biometrics systems. In concern of cost Authors observed that for cost benefit advantage in the initial implementation stages the finger print technology could be a better selection as compared to other available biometric technologies. In concern of security there are strong reasons that biometrics could be implemented because there are many secure authentication devices related to biometrics are available in the market that could secure data in best possible way. Authors observed that there is a need of legislations for biometrics as the security measures going to be much higher as compared to the traditional password systems. In concern of privacy authors observed that the organizations are demanding strong authentication and focus on suggesting biometrics because it could provide advantages to patients, healthcare staff and healthcare providers. / lineofaxis@gmail.com, ms24pk@gmail.com

Biometrics Technology is an authentication technology that identifies the individuals from their physical and behavioral characteristics. Despite the fact that biometrics technology provides robust authentication and enhanced security, it has not yet been implemented in many parts of the world due to certain issues i.e. human perceptions of the biometrics technology and cost factor, involved in the deployment of biometrics technology. As the biometrics technology involves identity management of individuals that’s why the humans perceptions of biometrics technology i.e. privacy concerns, security concerns and user acceptance issue play a very important role in the deployment of biometrics technology. There for the human perceptions and cost factor need to be considered before any deployment of biometrics technology. The aim of this thesis work is to study and analyze how the people’s perceptions and cost factor can be solved for the deployment of biometrics technology in Blekinge health care system. Literature study, interviews and survey are performed by authors for the identification and understanding of the human perceptions and cost factor. Based on these, solutions in form of guidelines to the issues involved in the biometrics technology deployment in Blekinge health care system Sweden are given. / Biometri Teknik är en autentisering teknik som identifierar individer från deras fysiska och beteendemässiga egenskaper. Trots att biometriska tekniken ger en robust autentisering och ökad säkerhet, har det ännu inte genomförts i många delar av världen på grund av vissa frågor som exempelvis mänskliga uppfattningar om biometriska tekniken och extra kostnader, som deltar i användningen av biometriska tekniken. Eftersom biometriska tekniken innebär identitetshantering av individer som är anledningen till att människor uppfattningar av biometriska tekniken dvs rör skyddet av privatlivet, säkerhetsfrågor och användarnas inställning fråga spelar en mycket viktig roll i spridningen av biometriska tekniken. Där för människors uppfattningar och kostnad faktor måste beaktas innan en eventuell utplacering av biometriska tekniken. Syftet med detta examensarbete är att studera och analysera hur människors uppfattning och kostnad faktor kan lösas för införande av biometri teknik i Blekinge hälsovårdssystemet. Litteraturstudie, intervjuer och undersökningar utförs av författare för identifiering och förståelse av människans förnimmelser och kostnad faktor. Baserat på dessa, är lösningar i form av riktlinjer för de frågor som berörs av biometri spridningen av tekniken i Blekinge sjukvårdssystem Sverige ges. / Falakzeb@ymail.com, Sajidnaseem.bth@gmail.com

Different security issues are a top subject around the world, especially since the terror threats seem to intensify. In the same time, the transport industry suffer from problems with smuggling and theft of valuable goods. One way to increase the security might be to have a verification system installed in commercial trucks, in order to assure that the driver is the proper one. This thesis has two purposes. One is to find appropriate methods for driver verification and build a prototype of a verification system which can be used for testing and further development. The other is to study how truck drivers perceive such a system and how their conception goes along with the growing demand for higher security. The present work is the result of a cooperation between an engineer and a cognitive scientist. The thesis focuses on the transport industry and was performed for Volvo Technology Corporation (VTEC), Gothenburg, Sweden. Eleven available verification methods were studied. To enable a well-based selection of methods to implement in the prototype, inquiries and interviews with truck drivers and haulage contractors were carried out to complement the theoretical study. One regular and three biometric verification methods were chosen for the test; fingerprint verification, face recognition, voice recognition and PIN verification. These methods were put together to a prototype system that was implemented in a truck simulator. A graphical user interface was developed in order to make the system user friendly. The prototype system was tested by 18 truck drivers. They were thoroughly interviewed before and after the test in order to retrieve their background, expectations and opinions as well as their perceptions and experiences of the test. Most of the test participants were positive to the prototype system. Even though they did not feel a need for it today they believed it to “be the future”. However, some participants felt uncomfortable with the system since they felt controlled by it. It became clear how important it is to have a system that respect the users’ privacy and to assure that the users are well informed about how the system is used. Some of the technology used for the verification system requires more development to fit in the automotive context, but it is considered to be possible to achieve a secure and robust system.

This thesis states that cracks and scars, referred to as Secondary Creases, in fingerprint images can be used as means for aiding and complementing fingerprint recognition, especially in cases where there is not enough clear data to use traditional methods such as minutiae based or correlation techniques. A Gabor filter bank is used to extract areas with linear patterns, where after the Hough Transform is used to identify secondary creases in a r, theta space. The methods proposed for Secondary Crease extraction works well, and provides information about what areas in an image contains usable linear pattern. Methods for comparison is however not as robust, and generates False Rejection Rate at 30% and False Acceptance Rate at 20% on the proposed dataset that consists of bad quality fingerprints. In short, our methods still makes it possible to make use of fingerprint images earlier considered unusable in fingerprint recognition systems.

Biometrics is a collection of methods for identifying and authorizing individuals with something they are, or do. It is considered to be one of the most secure technologies for security, both physical and logical. Security is something very important for organizations today, but yet there has been a low interest in investing in biometrics for security in Sweden. The purpose of this thesis has been to establish factors to why biometrics has not been implementedto any large extent, even though the technology has been around for long. We have evaluated the attitudes and views of individuals, as well as company leaders. Problem and questions with biometrics that often are discussed are privacy concerns, costs and ROI (return on investment) and user acceptance. Foremost there is a concern about the costs of implementing such a solution, although some claim that money can be saved by avoiding the costs in more traditional security methods, such as password administration and in security cards. There is a range of different technologies to use, such as facial-scan, voice-scan and the most mature and well known technique, finger-scan which has been around for a long time, especially for identifying criminals. All these techniques have there positive and negative sides, often measured in different levels of universality, uniqueness, permanence and collectability, e.g. eye-scan has a high uniqueness, facial-scan has a high universability and signature-scan has a low permanence level. We have collected our data through interviews with companies and suppliers of biometric systems as well as a questionnaire given to private persons. By interpreting the data in a qualitative way we have made our analysis and reached a conclusion. Our result shows that a cost related to biometric solutions is the largest reason why investments have been absent. This verifies the results of other authors, that the cost is the largest obstacle. Another important factor is that companies believe that biometrics is for organizations with a very high security need. Further our result show that individuals are positive towards biometrics. / Biometri är ett samlingsnamn för tekniker som identifierar och auktoriserar individer, antingen något de är eller gör. Biometri anses vara en av de säkraste teknologierna för säkerhet, både fysisk och logisk. Säkerhet är något som är mycket viktigt för organisationer i dagens läge, men än så länge så har investeringar i biometriska säkerhetslösningar i Sverige uteblivit. Syftet med denna magisteruppsats har varit att fastställa faktorer som bidrar till att biometriska lösningar inte har blivit implementerade i någon större utsträckning, trots att teknologin har funnits länge. Vi har utvärderat individers samt företagsledares attityder och synsätt angående biometri. Frågor och problem som ofta relateras till biometri är personligintegritet, kostnader och avkastning på investering samt användaracceptans. Framförallt är det osäkerheten kring kostnaderna av en implementering av en biometrisklösning. Det finns även de som hävdar att biometriska lösningar kan spara pengar jämfört med traditionella system därför att man undviker till exempel lösenordsadministration och passerkort. Det finns en rad olika tekniker att använda, exempelvis ansiktsavläsning, röstigenkänning och den mest mogna och kända tekniken, fingeravtrycksläsning som har existerat en längre tid, framförallt för att identifiera kriminella. Det finns positiva och negativa sidor med alla dessa tekniker, de mäts oftast i olika nivåer av hur många som kan använda det, hur unikt biometrin är, beständighet och hur biometrin samlas in. Genom intervjuer med företag och leverantörer av biometriska lösningar samt en utdelad enkät till privat personer har vi samlat in data. Vi har sedan tolkat data kvalitativt och utfört vår analys och slutligen kommit fram till ett resultat. Vårt resultat har visat att kostnader relaterade till biometriska system är det största skälet till varför investeringar har uteblivit. Detta bekräftar vad många andra författare tidigare har funnit, att kostnaderna är det största hindret. En annan viktig faktor är att företag anser att biometri är något för verksamheter som kräver den allra högsta säkerheten. Vidare har individer en positiv inställning till biometri.

Abstract Because of its natural and non-intrusive interaction, identity verification and recognition using facial information is among the most active areas in computer vision research. Unfortunately, it has been shown that conventional 2D face recognition techniques are vulnerable to spoofing attacks, where a person tries to masquerade as another one by falsifying biometric data and thereby gaining an illegitimate advantage. This thesis explores different directions for software-based face anti-spoofing. The proposed approaches are divided into two categories: first, low-level feature descriptors are applied for describing the static and dynamic characteristic differences between genuine faces and fake ones in general, and second, complementary attack-specific countermeasures are investigated in order to overcome the limitations of generic spoof detection schemes. The static face representation is based on a set of well-known feature descriptors, including local binary patterns, Gabor wavelet features and histogram of oriented gradients. The key idea is to capture the differences in quality, light reflection and shading by analysing the texture and gradient structure of the input face images. The approach is then extended to the spatiotemporal domain when both facial appearance and dynamics are exploited for spoof detection using local binary patterns from three orthogonal planes. It is reasonable to assume that no generic spoof detection scheme is able to detect all known, let alone unseen, attacks scenarios. In order to find out well-generalizing countermeasures, the problem of anti-spoofing is broken into two attack-specific sub-problems based on whether the spoofing medium can be detected in the provided view or not. The spoofing medium detection is performed by describing the discontinuities in the gradient structures around the detected face. If the display medium is concealed outside the view, a combination of face and background motion correlation measurement and texture analysis is applied. Furthermore, an open-source anti-spoofing fusion framework is introduced and its system-level performance is investigated more closely in order to gain insight on how to combine different anti-spoofing modules. The proposed spoof detection schemes are evaluated on the latest benchmark datasets. The main findings of the experiments are discussed in the thesis. / Tiivistelmä Kasvokuvaan perustuvan henkilöllisyyden tunnistamisen etuja ovat luonnollinen vuorovaikutus ja etätunnistus, minkä takia aihe on ollut erittäin aktiivinen tutkimusalue konenäön tutkimuksessa. Valitettavasti tavanomaiset kasvontunnistustekniikat ovat osoittautuneet haavoittuvaisiksi hyökkäyksille, joissa kameralle esitetään jäljennös kohdehenkilön kasvoista positiivisen tunnistuksen toivossa. Tässä väitöskirjassa tutkitaan erilaisia ohjelmistopohjaisia ratkaisuja keinotekoisten kasvojen ilmaisuun petkuttamisen estämiseksi. Työn ensimmäisessä osassa käytetään erilaisia matalan tason piirteitä kuvaamaan aitojen ja keinotekoisten kasvojen luontaisia staattisia ja dynaamisia eroavaisuuksia. Työn toisessa osassa esitetään toisiaan täydentäviä hyökkäystyyppikohtaisia vastakeinoja, jotta yleispätevien menetelmien puutteet voitaisiin ratkaista ongelmaa rajaamalla. Kasvojen staattisten ominaisuuksien esitys perustuu yleisesti tunnettuihin matalan tason piirteisiin, kuten paikallisiin binäärikuvioihin, Gabor-tekstuureihin ja suunnattujen gradienttien histogrammeihin. Pääajatuksena on kuvata aitojen ja keinotekoisten kasvojen laadun, heijastumisen ja varjostumisen eroavaisuuksia tekstuuria ja gradienttirakenteita analysoimalla. Lähestymistapaa laajennetaan myös tila-aika-avaruuteen, jolloin hyödynnetään samanaikaisesti sekä kasvojen ulkonäköä ja dynamiikkaa irroittamalla paikallisia binäärikuvioita tila-aika-avaruuden kolmelta ortogonaaliselta tasolta. Voidaan olettaa, ettei ole olemassa yksittäistä yleispätevää vastakeinoa, joka kykenee ilmaisemaan jokaisen tunnetun hyökkäystyypin, saati tuntemattoman. Näin ollen työssä keskitytään tarkemmin kahteen hyökkäystilanteeseen. Ensimmäisessä tapauksessa huijausapuvälineen reunoja ilmaistaan analysoimalla gradienttirakenteiden epäjatkuvuuksia havaittujen kasvojen ympäristössä. Jos apuvälineen reunat on piilotettu kameran näkymän ulkopuolelle, petkuttamisen ilmaisu toteutetaan yhdistämällä kasvojen ja taustan liikkeen korrelaation mittausta ja kasvojen tekstuurianalyysiä. Lisäksi työssä esitellään vastakeinojen yhdistämiseen avoimen lähdekoodin ohjelmisto, jonka avulla tutkitaan lähemmin menetelmien fuusion vaikutuksia. Tutkimuksessa esitetyt menetelmät on kokeellisesti vahvistettu alan viimeisimmillä julkisesti saatavilla olevilla tietokannoilla. Tässä väitöskirjassa käydään läpi kokeiden päähavainnot.

Abstract In this work, statistical machine learning and pattern recognition methods were developed and applied to sensor-based human context recognition. More precisely, we concentrated on an effective discriminative learning framework, where input-output mapping is learned directly from a labeled dataset. Non-parametric discriminative classification and regression models based on kernel methods were applied. They include support vector machines (SVM) and Gaussian processes (GP), which play a central role in modern statistical machine learning. Based on these established models, we propose various extensions for handling structured data that usually arise from real-life applications, for example, in a field of context-aware computing. We applied both SVM and GP techniques to handle data with multiple classes in a structured multi-sensor domain. Moreover, a framework for combining data from several sources in this setting was developed using multiple classifiers and fusion rules, where kernel methods are used as base classifiers. We developed two novel methods for handling sequential input and output data. For sequential time-series data, a novel kernel based on graphical presentation, called a weighted walk-based graph kernel (WWGK), is introduced. For sequential output labels, discriminative temporal smoothing (DTS) is proposed. Again, the proposed algorithms are modular, so different kernel classifiers can be used as base models. Finally, we propose a group of techniques based on Gaussian process regression (GPR) and particle filtering (PF) to learn to track multiple targets. We applied the proposed methodology to three different human-motion-based context recognition applications: person identification, person tracking, and activity recognition, where floor (pressure-sensitive and binary switch) and wearable acceleration sensors are used to measure human motion and gait during walking and other activities. Furthermore, we extracted a useful set of specific high-level features from raw sensor measurements based on time, frequency, and spatial domains for each application. As a result, we developed practical extensions to kernel-based discriminative learning to handle many kinds of structured data applied to human context recognition. / Tiivistelmä Tässä työssä kehitettiin ja sovellettiin tilastollisen koneoppimisen ja hahmontunnistuksen menetelmiä anturipohjaiseen ihmiseen liittyvän tilannetiedon tunnistamiseen. Esitetyt menetelmät kuuluvat erottelevan oppimisen viitekehykseen, jossa ennustemalli sisääntulomuuttujien ja vastemuuttujan välille voidaan oppia suoraan tunnetuilla vastemuuttujilla nimetystä aineistosta. Parametrittomien erottelevien mallien oppimiseen käytettiin ydinmenetelmiä kuten tukivektorikoneita (SVM) ja Gaussin prosesseja (GP), joita voidaan pitää yhtenä modernin tilastollisen koneoppimisen tärkeimmistä menetelmistä. Työssä kehitettiin näihin menetelmiin liittyviä laajennuksia, joiden avulla rakenteellista aineistoa voidaan mallittaa paremmin reaalimaailman sovelluksissa, esimerkiksi tilannetietoisen laskennan sovellusalueella. Tutkimuksessa sovellettiin SVM- ja GP-menetelmiä moniluokkaisiin luokitteluongelmiin rakenteellisen monianturitiedon mallituksessa. Useiden tietolähteiden käsittelyyn esitetään menettely, joka yhdistää useat opetetut luokittelijat päätöstason säännöillä lopulliseksi malliksi. Tämän lisäksi aikasarjatiedon käsittelyyn kehitettiin uusi graafiesitykseen perustuva ydinfunktio sekä menettely sekventiaalisten luokkavastemuuttujien käsittelyyn. Nämä voidaan liittää modulaarisesti ydinmenetelmiin perustuviin erotteleviin luokittelijoihin. Lopuksi esitetään tekniikoita usean liikkuvan kohteen seuraamiseen. Menetelmät perustuvat anturitiedosta oppivaan GP-regressiomalliin ja partikkelisuodattimeen. Työssä esitettyjä menetelmiä sovellettiin kolmessa ihmisen liikkeisiin liittyvässä tilannetiedon tunnistussovelluksessa: henkilön biometrinen tunnistaminen, henkilöiden seuraaminen sekä aktiviteettien tunnistaminen. Näissä sovelluksissa henkilön asentoa, liikkeitä ja astuntaa kävelyn ja muiden aktiviteettien aikana mitattiin kahdella erilaisella paineherkällä lattia-anturilla sekä puettavilla kiihtyvyysantureilla. Tunnistusmenetelmien laajennuksien lisäksi jokaisessa sovelluksessa kehitettiin menetelmiä signaalin segmentointiin ja kuvaavien piirteiden irroittamiseen matalantason anturitiedosta. Tutkimuksen tuloksena saatiin parannuksia erottelevien mallien oppimiseen rakenteellisesta anturitiedosta sekä erityisesti uusia menettelyjä tilannetiedon tunnistamiseen.