This thesis introduces the concept of a person recognition system for use on an integrated autonomous surveillance camera. Developed to enable generic surveillance tasks without the need for complex setup procedures nor operator assistance, this is achieved through the novel use of a simple dynamic noise reduction and object detection algorithm requiring no previous knowledge of the installation environment and without any need to train the system to its installation. The combination of this initial processing stage with a novel hybrid neural network structure composed of a SOM mapper and an MLP classifier using a combination of common and individual input data lines has enabled the development of a reliable detection process, capable of dealing with both noisy environments and partial occlusion of valid targets. With a final correct classification rate of 94% on a single image analysis, this provides a huge step forwards as compared to the reported 97% failure rate of standard camera surveillance systems.

This thesis reports research conducted into a character recognition system suitable for use in a reading aid for the blind. A brief review of blind reading aids is given, showing the need for a device which is cheap, simple and effective. The structure of a proposed reading aid fulfilling these needs is outlined, with a list of the desired characteristics of each of its subsystems. The remainder of the thesis is concerned with research into just two of these subsystems: the input device and the character recognizer. A detailed review of pattern recognition by the n-tuple method is presented, followed by a description of the experimental techniques used in obtaining real data from a camera system, and in simulating various recognizer structures. The camera system and computer programs developed specifically for the research are described in detail. Several series of experiments are reported, concerned mainly with investigating problems associated directly with the blind reading aid, namely accommodation of multifont printed text and of the tracking errors inherent in data from a hand-held probe. A further series of experiments, aimed at improving the performance of the recognizer within fixed size constraints, i. e., optimisation, has a wider field of application. Finally suggestions are made as to how the recognizer might be implemented in a reading aid, using RAMs, ROMs, or PLAs as the main storage elements.

Masters of Science / This research investigates the creation of an audio-shape recognition system that is able to interpret a user’s drawn audio shapes—fundamental shapes, digits and/or letters— on a given surface such as a table-top using a generic stylus such as the back of a pen. The system aims to make use of one, two or three Piezo microphones, as required, to capture the sound of the audio gestures, and a combination of the Mel-Frequency Cepstral Coefficients (MFCC) feature descriptor and Support Vector Machines (SVMs) to recognise audio shapes. The novelty of the system is in the use of piezo microphones which are low cost, light-weight and portable, and the main investigation is around determining whether these microphones are able to provide sufficiently rich information to recognise the audio shapes mentioned in such a framework.

The rapid advance of information technology has allowed for a rise in the use of biometric markers to automatically track the identity of individuals. Iris biometrics has emerged as one of the most reliable and accurate systems when dealing with cooperating subjects, however, challenges arise when attempting to minimize the amount of intrusion when examining subjects. Allowing for more flexibility in data capture settings will introduce differences in the iris texture due to changes in ambient light, which may negatively impact recognition results. This research examines the feasibility of using 3D software to synthetically dilate the pupils of existing iris images to more closely match the size of a target image. Methods are developed first to evaluate the compatibility of synthetic images with iris identification software, and then to examine what specific areas of the iris texture differ between synthetic and real images. Results show synthetic images are found to be compatible with the recognition process and have the potential to improve performance.

The analysis carried out in this thesis is made from two different points of view, the qualitative and the quantitative, by using the case study of Malmö intermodal terminal. The first analysis is focused on how the intermodal terminals works and which elements of it interact and how, in order to achieve the purpose of the terminal, and how the Intelligent Video Gate is able to affect in any way to this functioning, mainly in a positive way that allows the better functioning of the terminal.From the quantitative point of view what is carried out is a timing and economic analysis of the Malmö Intermodal Terminal, which is based on the information obtained from the qualitative analysis and from the data provided by the terminal operators that allow to make different simulations to compare the effect of the Intelligent Video Gate implementation in this specific terminal, and that could be extended to similar intermodal terminals located in regions with similar labour conditions and that as the European Union have a huge standardized freight system.Finally, what is stated with the provided data, despite not allowing to make the most complex and representative simulation, is that the aim of the Intelligent Video Gate is reached successfully with a great improvement of the efficiency what allows to ensure with quite certainty that the system implementation is recommended in this kind of terminals.

Facial Recognition Systems är något som har blivit populärt de senaste åren, speciellt efter den 11 september 2001. Möjligheten att kunna över-vaka personer som rör sig i olika miljöer har varit av intresse för bland annat regeringar, till exempel USA:s regering. Eftersom det finns mängder med olika typer av undersökningar och alla försöker göra så bra matchningar som möjligt av personer mot databaser, fast de utförs på olika sätt, är det intressant att se om det finns någon parameter som har en större påverkan på resultaten, oavsett om undersökningen görs med 2D, 3D eller en kombination av metoderna. Det finns många olika faktorer och parametrar som påverkar matchningsprocenten därför skall denna littera-turstudie försöka lokalisera och se om det finns någon parameter som har en större påverkan på matchningsprocenten. Det visar sig att två parametrar har en större påverkan än övriga parametrar. De är antalet bilder av varje objekt som finns att matcha i databasen och kvaliteten på indata vilket innebär kontrast och upplösning samt hur kompletta ansiktena är. / Facial Recognition Systems is something that has become popular in recent years, especially after 11 September 2001. The ability to monitor people that are moving in different environments has been of interest to particular governments, for instance the US government. Since there are a lot of different types of surveys, though performed in different ways, all trying to do the best matches of people to databases as possible, it is interesting to see if there is any parameter that has a major impact on the result. Whether the survey is done with 2D, 3D or a combination of methods, there are many different factors and parameters that affect the matching percentage. Therefore this study tries to locate and see if there is any parameter that has a greater impact on the matching percentage. It appears that two of the parameters have a greater effect on the result, than the others. These are the number of images of a test subject in the database and the quality of the input data. The quality is defined by contrast and resolution as well as how complete the faces are.

Questions of speaker’s recognition by voice are investigated in this dissertation. Speaker recognition systems, their evolution, problems of recognition, systems of features, questions of speaker modeling and matching used in text-independent and text-dependent speaker recognition are considered too. The text-independent speaker recognition system has been developed during this work. The Gaussian mixture model approach was used for speaker modeling and pattern matching. The automatic method for voice activity detection was proposed. This method is fast and does not require any additional actions from the user, such as indicating patterns of the speech signal and noise. The system of the features was proposed. This system consists of parameters of excitation source (glottal) and parameters of the vocal tract. The fundamental frequency was taken as an excitation source parameter and four formants with three antiformants were taken as parameters of the vocal tract. In order to equate dispersions of the formants and antiformants we propose to use them in mel-frequency scale. The standard mel-frequency cepstral coefficients (MFCC) for comparison of the results were implemented in the recognition system too. These features make baseline in speech and speaker recognition. The experiments of speaker recognition have shown that our proposed system of features outperformed standard mel-frequency cepstral coefficients. The equal error rate (EER) was equal to 5.17% using proposed... [to full text] / Disertacijoje nagrinėjami kalbančiojo atpažinimo pagal balsą klausimai. Aptartos kalbančiojo atpažinimo sistemos, jų raida, atpažinimo problemos, požymių sistemos įvairovė bei kalbančiojo modeliavimo ir požymių palyginimo metodai, naudojami nuo ištarto teksto nepriklausomame bei priklausomame kalbančiojo atpažinime. Darbo metu sukurta nuo ištarto teksto nepriklausanti kalbančiojo atpažinimo sistema. Kalbėtojų modelių kūrimui ir požymių palyginimui buvo panaudoti Gauso mišinių modeliai. Pasiūlytas automatinis vokalizuotų garsų išrinkimo (segmentavimo) metodas. Šis metodas yra greitai veikiantis ir nereikalaujantis iš vartotojo jokių papildomų veiksmų, tokių kaip kalbos signalo ir triukšmo pavyzdžių nurodymas. Pasiūlyta požymių vektorių sistema, susidedanti iš žadinimo signalo bei balso trakto parametrų. Kaip žadinimo signalo parametras, panaudotas žadinimo signalo pagrindinis dažnis, kaip balso trakto parametrai, panaudotos keturios formantės bei trys antiformantės. Siekiant suvienodinti žemesnių bei aukštesnių formančių ir antiformančių dispersijas, jas pasiūlėme skaičiuoti melų skalėje. Rezultatų palyginimui sistemoje buvo realizuoti standartiniai požymiai, naudojami kalbos bei asmens atpažinime – melų skalės kepstro koeficientai (MSKK). Atlikti kalbančiojo atpažinimo eksperimentai parodė, kad panaudojus pasiūlytą požymių sistemą buvo gauti geresni atpažinimo rezultatai, nei panaudojus standartinius požymius (MSKK). Gautas lygių klaidų lygis, panaudojant pasiūlytą požymių... [toliau žr. visą tekstą]

The tremendous increase in usage and complexity of modern communication and network systems connected to the Internet, places demands upon security management to protect organisations' sensitive data and resources from malicious intrusion. Malicious attacks by intruders and hackers exploit flaws and weakness points in deployed systems through several sophisticated techniques that cannot be prevented by traditional measures, such as user authentication, access controls and firewalls. Consequently, automated detection and timely response systems are urgently needed to detect abnormal activities by monitoring network traffic and system events. Network Intrusion Detection Systems (NIDS) and Network Intrusion Prevention Systems (NIPS) are technologies that inspect traffic and diagnose system behaviour to provide improved attack protection. The current implementation of intrusion detection systems (commercial and open-source) lacks the scalability to support the massive increase in network speed, the emergence of new protocols and services. Multi-giga networks have become a standard installation posing the NIDS to be susceptible to resource exhaustion attacks. The research focuses on two distinct problems for the NIDS: missing alerts due to packet loss as a result of NIDS performance limitations; and the huge volumes of generated alerts by the NIDS overwhelming the security analyst which makes event observation tedious. A methodology for analysing alerts using a proposed framework for alert correlation has been presented to provide the security operator with a global view of the security perspective. Missed alerts are recovered implicitly using a contextual technique to detect multi-stage attack scenarios. This is based on the assumption that the most serious intrusions consist of relevant steps that temporally ordered. The pre- and post- condition approach is used to identify the logical relations among low level alerts. The alerts are aggregated, verified using vulnerability modelling, and correlated to construct multi-stage attacks. A number of algorithms have been proposed in this research to support the functionality of our framework including: alert correlation, alert aggregation and graph reduction. These algorithms have been implemented in a tool called Multi-stage Attack Recognition System (MARS) consisting of a collection of integrated components. The system has been evaluated using a series of experiments and using different data sets i.e. publicly available datasets and data sets collected using real-life experiments. The results show that our approach can effectively detect multi-stage attacks. The false positive rates are reduced due to implementation of the vulnerability and target host information.

薪俸制度的良窳，不僅影響優秀人才的延攬和保持，更與人員士氣和績效的提升有密切相關。因此，美國聯邦政府師法民營企業實施績效薪俸(Pay for Performance)制度之成效，於一九七八年文官改革法，針對一般俸表第十三至十五職等的管理人員與監督人員創設功績俸制度(MPS)，摒棄過去自動晉俸的調整薪俸方式，改依工作績效決定加薪數額，期能激勵人員士氣與績效，進而提昇組織生產力。 　　本論文旨在探討美國聯邦政府施行功績俸制度之情形，採取文獻分析法，首先檢視支持與反對功績俸制度的相關理論，其次針對美國聯邦政府實施功績俸制度之過程和內容，作一有系統的介紹，並綜理其施行結果，檢討其成敗關鍵，進而說明有效實施該制的要件。 　　本研究結果發現，功績俸制度和其後在一九八四年改革之績效管理與獎賞制度(PMRS)所強調的「依功論酬」概念頗佳，也得到員工的支持，但是大部分的實際研究均指出，兩制並未能達成改進組織生產力的目標。績效評估的困難與基金不足的問題，乃是影響制度成敗的兩大關鍵因素。值得一提的是，儘管一連串的功績俸計畫成效不佳，但是聯邦政府仍不願放棄該一理念，反而積極檢討將該制擴大的可能性。 　　最後，除了對功績俸制度提出改進建議外，並探討該制在我國政府機關的適用性問題，以及歸納政府機關施行績效薪俸制度之原則以供參考。

Visual recognition research develops algorithms and representations to autonomously recognize visual entities such as objects, actions, and attributes. The traditional protocol involves manually collecting training image examples, annotating them in specific ways, and then learning models to explain the annotated examples. However, this is a rather limited way to transfer human knowledge to visual recognition systems, particularly considering the immense number of visual concepts that are to be learned. I propose new forms of active learning that facilitate large-scale transfer of human knowledge to visual recognition systems in a cost-effective way. The approach is cost-effective in the sense that the division of labor between the machine learner and the human annotators respects any cues regarding which annotations would be easy (or hard) for either party to provide. The approach is large-scale in that it can deal with a large number of annotation types, multiple human annotators, and huge pools of unlabeled data. In particular, I consider three important aspects of the problem: (1) cost-sensitive multi-level active learning, where the expected informativeness of any candidate image annotation is weighed against the predicted cost of obtaining it in order to choose the best annotation at every iteration. (2) budgeted batch active learning, a novel active learning setting that perfectly suits automatic learning from crowd-sourcing services where there are multiple annotators and each annotation task may vary in difficulty. (3) sub-linear time active learning, where one needs to retrieve those points that are most informative to a classifier in time that is sub-linear in the number of unlabeled examples, i.e., without having to exhaustively scan the entire collection. Using the proposed solutions for each aspect, I then demonstrate a complete end-to-end active learning system for scalable, autonomous, online learning of object detectors. The approach provides state-of-the-art recognition and detection results, while using minimal total manual effort. Overall, my work enables recognition systems that continuously improve their knowledge of the world by learning to ask the right questions of human supervisors. / text