Social anxiety is a disorder where people fear social interactions and is associated with physiological changes. Eye tracking studies have shown that people with social anxiety spent more time gazing at emotional faces presented on a computer screen and spent more time gazing at the eye region. There has been limited studies on tracking eye gaze in a real-life setting interacting with another person. We used a wearable eye tracker during a brief one-on-one interview about participants’ challenges faced at work or school. Along with self-report psychological measures about social anxiety and shyness, we also measured participants’ salivary cortisol as a metric for physiological stress. We hypothesized that socially anxious individuals would have higher cortisol levels and spent more time gazing at the face. However, there was no change in cortisol levels before and after the interview. In addition, socially anxious individuals had lower cortisol levels than less anxious people. Furthermore, the time spent fixating on the region of interest (ROI), which was the face, was not correlated with social anxiety, anxiety or shyness. Paradoxically, the more socially anxious participants seem to have had a lower physiological stress response than less socially anxious participants.

Video is the most demanding modality from the viewpoints of bandwidth, computational complexity, and resolution. Thus, there has been limited progress in the field of mobile video technology. In the research, the focus is on elastic wireless video technology, and its adaptation to diagnostic application requirements in real-time clinical assessment. It is important and timely to apply wireless video technology to real-time remote diagnosis of emergent medical events. This premise comes from initial successes in telehealth based on wired networks. The enablement of mobility (for the physician and/or the patient) by wireless communication will be a next major step, but this advance will depend on definitive and compelling demonstrations of reliability. Thus, an important goal of the research is to develop a complete methodology that will be embraced by physicians. Acute pediatric asthma has been identified as a domain where this new capability will be highly welcome. The research uses flexible and interactive algorithms for Region-of-Interest (ROI) processing. ROI processing is a useful approach to achieve the optimal balance in the quality-bandwidth tradeoff characteristic of visual communication services. The notion of ROI has been traditionally used mostly for foreground-background separation in scene rendering and manipulation, and only more recently for variably quality compression. Even when the latter goal is considered, quality criteria have been ad-hoc and at best useful for video conferencing, given that the medical domain has its own fidelity criteria. The research thus focuses on the design of an elastic ROI-based compression paradigm with medical diagnosis as a central criterion. The research describes the methodology to achieve elasticity through rate control algorithms at the encoder. An elastic non-parametric approach is proposed that uses a priori user-specified video quality information, quantifies this information, and incorporates this into the encoder in the form of region-quality mappings. This method is compared to a parametric bit allocation approach that is based on region-features and a set of tuning weights. A number of videos of actual patients were filmed and used as the video database for the developed algorithms. In testing the elastic non-parametric and parametric algorithms, both objective measures in the form of Peak Signal to Noise Ratio (PSNR), and subjective evaluations were used. Thus, in this work, the focus is on domain relevance of the algorithms developed, as opposed to network related issues such as packet losses. This is justified in that these may have broader value with other applications, and continuation of this work will include realistic network conditions. To summarize, the research shows the usefulness of ROI processing as a means of achieving a gain (in a bits per pixel sense) over uniform compression at the same bitrate. It also shows how quantifying a notion of functionally lossless video quality diagnostically lossless video quality in a video-based telehealth system, in a bits per pixel sense is useful from an applications and bitrate perspective. Finally, a combination of these two concepts is advantageous i.e. diagnostically lossless ROI video quality is achievable over bitrate limited channels.

Τα τελευταία χρόνια, η δημοτικότητα του διαδικτύου αυξάνεται ολοένα και περισσότερο και σε συνδυασμό με την κινηματογραφική βιομηχανία που ανθίζει με γρήγορους ρυθμούς , έχει σαν αποτέλεσμα έναν τεράστιο αριθμό βίντεο κοινής χρήσης στο διαδίκτυο και μια πληθώρα κινηματογραφικών ταινιών, στα οποία έχει άμεση πρόσβαση μεγάλη μερίδα του πληθυσμού, συμπεριλαμβανομένων και διάφορων ευαίσθητων κοινωνικών ομάδων, παραδείγματος χάρη παιδιά και εφήβους. Η προστασία τέτοιων ατόμων αλλά και η επιθυμία γνώσης του περιεχομένου ενός βίντεο δημιούργησε την αναγκαιότητα ανάπτυξης αποτελεσματικών, αυτόματων ανιχνευτών βίας.Στην παρούσα διπλωματική παρουσιάζονται οι μέθοδοι που έχουν προταθεί στο συγκεκριμένο πεδίο. Στην συνέχεια, υιοθετείται μια εκ των μεθόδων και αναπτύσσεται αλγόριθμος, με σκοπό τη μελέτη της απόδοσης του. / In recent years, the popularity of the internet growing more and more.This results a huge number of video sharing on the internet and a plethora of films. A large portion of population has direct access in such videos,including sensitive and different social groups , for example children and adolescents . The protection of such persons and the desire knowing the content of a video, created the necessity to develop efficient , automated violence detectors.In this dissertation we present methods that have been proposed in this field . Then , we have adopted one of the methods and we have developed an algorithm in order to study its accuracy.

A novel method is presented for analyzing fMRI data, which relies on probabilistic estimates of microanatomically defined regions in individual fMRI volunteers. Postmortem structural and cytoarchitectonic information from the Julich/Dusseldorf group in Germany is aligned to the high-resolution structural MR images of functional MRI volunteers. This is achieved using nonlinear registration, which is applied only to the region of interest. The registered postmortem datasets are then combined into probability maps for microanatomically defined regions that are tailored to the anatomy of individual fMRI volunteers. These are then used as weighted spatial filters on functional MR data. In this thesis, three regions of the primary auditory cortex (located on Heschl's gyrus) have been targeted, and the analysis method is used to explore how these three areas respond to different kinds of sound. Regions Te1.0 and Te1.2 both demonstrate pitch sensitivity, consistent with published observations of the functional response of homologous regions in nonhuman primates. Area Te1.1 displayed sensitivity to both noise and pitch, consistent with the theory that it is homologous with the microanatomically similar area CM in nonhuman primates. Furthermore, the custom probability maps are much less diffuse and anatomically more precise than previous versions generated using the same postmortem data, and therefore permit a more sensitive and anatomically precise analysis of functional activity. This method could be applied to any other microanatomically defined region that has been characterized in the Julich postmortem dataset. / Thesis (Master, Computing) -- Queen's University, 2008-09-26 19:50:54.582

No description available.

As video capturing devices become more ubiquitous from surveillance cameras to smart phones, the demand of automated video analysis is increasing as never before. One obstacle in this process is to efficiently locate where a human operator’s attention should be, and another is to determine the specific types of activities or actions without ambiguity. It is the special interest of this dissertation to locate spatial and temporal regions of interest in videos and to develop a better action representation for video-based activity analysis. This dissertation follows the scheme of “locating then recognizing” activities of interest in videos, i.e., locations of potentially interesting activities are estimated before performing in-depth analysis. Theoretical properties of regions of interest in videos are first exploited, based on which a unifying framework is proposed to locate both spatial and temporal regions of interest with the same settings of parameters. The approach estimates the distribution of motion based on 3D structure tensors, and locates regions of interest according to persistent occurrences of low probability. Two contributions are further made to better represent the actions. The first is to construct a unifying model of spatio-temporal relationships between reusable mid-level actions which bridge low-level pixels and high-level activities. Dense trajectories are clustered to construct mid-level actionlets, and the temporal relationships between actionlets are modeled as Action Graphs based on Allen interval predicates. The second is an effort for a novel and efficient representation of action graphs based on a sparse coding framework. Action graphs are first represented using Laplacian matrices and then decomposed as a linear combination of primitive dictionary items following sparse coding scheme. The optimization is eventually formulated and solved as a determinant maximization problem, and 1-nearest neighbor is used for action classification. The experiments have shown better results than existing approaches for regions-of-interest detection and action recognition.

This thesis focuses on creating an application software for monitoring and analyzing changes in body surface temperature before and after exercise test by the means of contactless thermography. Initial, theoretical part of this thesis talks about principles and possible scenarios, in which one could employ infrared thermography followed by physiology of the exercise load and concludes with an observation of the changes in body surface temperature while focusing on the actions that take place during the exercise test. Practical part of the thesis comprises of the exercise test and building of the application software for the analysis of the acquired thermographic visual output. The application software offers 2 kinds of analyses: general and under-load, while it aims to provide evaluation in both cases based on user-selected areas of interest. Last chapter of this thesis includes final output with commentary and eventual usage of the under-load experiment and application software in practical use.

Die Grundlage für die Beurteilung von Szintigrammen im Bereich der Gliedmaße stellt der visuelle Vergleich mit der kontralateralen Seite dar. Eine Objektivierung anhand quantitativer Methoden ist daher, insbesondere bei Pferden mit bilateraler Lahmheit, von Bedeutung. Als Maß für den Anreicherungsgrad des Radiopharmakons in einem Areal werden sogenannte Speicherquotienten errechnet. In der Literatur finden sich für Speicherungen im Bereich des Strahlbeins nur wenig vergleichbare Werte, da die Auswahl der für die Bildung von Speicherquotienten nötigen Referenzareale nicht einheitlich erfolgt. Über die Auswertung im Bereich der Insertion der tiefen Beugesehne am Hufbein existieren kaum Angaben. Das Ziel der vorliegenden Arbeit ist es, Speicherquotienten für Anreicherungen im Bereich des Strahlbeins und der Insertion der tiefen Beugesehne am Hufbein zu ermitteln. Von Interesse ist vor allem die Beurteilung der Diskriminanz zwischen Patienten- und Kontrollgruppe in Abhängigkeit unterschiedlich gewählter Referenzareale. Weiterhin stellt sich die Frage, inwieweit Assoziationen zwischen Speicherquotienten und Lahmheitsgrad bzw. röntgenologischen Veränderungen bestehen. / The use of bone scan images of the distal equine limb allows for the comparison of differences of radiopharmaceutical uptake between the contralateral limbs for diagnostic purposes. This type of quantitation is particularly valuable in horses which demonstrate bilateral lameness. In order to quantitate and compare the density of radiopharmaceutical in each of the limbs, one can compare the uptake from the so-called region of interest to that of a region of reference in the same leg. This method overcomes problems incurred in using values obtained from the literature because it is difficult to compare ratios of uptake when the choice of the reference areas are not the same. In addition, comparable values which are available are not well standardized. Specifically, the area of insertion of the deep digital flexor tendon (DDFT) to the coffin bone is hardly ever discussed. The primary objective of this study was to assess the ratios for increased uptake in the navicular bone and the area of insertion of the deep digital flexor tendon to the coffin bone. Of particular interest was the ability to discriminate between the diseased group and the control group which depended significantly on the reference points used. An additional question was whether or not any associations existed between ratios of uptake, degree of lameness or presence of radiolographic changes.

Volvo Construction Equipment (CE) are facing the challenge of vibrations in their wheel loaders that generate disruptive noise and impact the driver's experience. These vibrations have been linked to the contact surface between the crown wheel and pinion gear in the vehicles drive-axles. In response, this thesis was created to develop an Artificial Intelligence (AI) system, which can identify outliers in a dataset containing images of the contact surfaces between the crown wheel and pinion gear. However, the dataset exhibits variations in image sharpness, exposure and centering of the crown wheel, which hinders its suitability for machine vision tasks. The varying quality of the images poses the challenge of accurately extracting relevant features required to analyze the images through machine learning algorithms. This research aims to address these challenges by investigating two research questions. (1) what method can be employed to extract the Region of Interest (ROI) in images of crown wheels? And (2) which method is suitable for detection of outliers within the ROI? To find answers to these questions, a literature study was conducted leading up to the implementation of two architectures: You Only Look Once (YOLO) v5 Oriented Bounding Boxes (OBB) and a Hybrid Autoencoder (BAE). Visual evaluation of the results showed promising outcomes particularly for the extraction of ROIs, where the relevant areas were accurately identified despite the large variations in image quality. The BAE successfully identified outliers that deviated from the majority, however, the results of the model were influenced by the differences in image quality, rather than the geometrical shape of the contact patterns. These findings suggest that using the same feature extraction method on a higher-quality dataset or employing a more robust segmentation method, could increase the likelihood of identifying the contact patterns responsible for the vibrations.

Panoramic virtual reality is an emerging technology that has recently gained the attention of both the research community and regular consumers. It allows the users to immerse themselves in omnidirectional videos with the help of a virtual reality headset : thanks to an increasing amount of affordable head-mounted-displays, any recent smartphone can offer a decent panoramic virtual reality experience. However since omnidirectional videos are videos with a large field-of-view that covers the entire sphere around the camera, they require large resolutions and thus high bitrates. This master degree project conducted at RE’FLEKT GmbH is an exploratory work that seeks to reduce the panoramic video bitrate. Because of the nature of omnidirectional videos, the user can only see a subpart of each video frame, and thus some zones of the video can attract more attention than others. The purpose of this study is to introduce the concept of region-of-interest encoding in panoramic VR. The main contribution is a method to encode panoramic videos in an H.264 video format stream with a space-variant level of details depending on the zones that attract the most the viewers’ interest. First, the region-of-interest are detected through a head-tracking module combined with a Gaussian attention model. Then, the reference video is encoded with the open source x264 encoder, with a quantization step adjusted to the region-of-interest information. The International Telecommunications Union standard subjective tests show that this method can perform better than classic H.264 encoding only in specific cases. / Panoramisk virtuell verklighet (VR) är en kommande teknik som nyligen har mött intresse från forskarsamhället och vanliga konsumenter. Det gör det möjligt för användarna att fördjupa sig i videor upptagna från flera riktningar, med hjälp av ett VR-headset : tack vare ett växande antal billiga och huvudburna bildskärmar, erbjuder alla nya smarttelefoner en passande panoramisk VR-erfarenhet. Men på grund av den breda synvinkeln i flerriktade media behöver videor med 360 graders synfält stor upplösning och därför höga bithastigheter. Detta masterexamensarbete som utförts på RE’FLEKT GmbH är ett utforskande arbete som strävar efter att reducera panoramabildens bithastighet. I flerriktade videoklipp kan användaren bara se en del av varje bildruta, härigenom får somliga zoner mer uppmärksamhet än andra. Syftet med denna studie är att introducera begreppet region-av-intresse (ROI) kodning i panoramisk VR. Huvudbidraget är en metod för att koda panoramisk video i en H.264-ström med en varierande nivå av detaljer som beror på de zoner som får mest av tittarnas intresse. Först detekteras ROI genom en huvudspårningsmodul kombinerad med en gaussisk uppmärksamhetsmodell. Därefter kodas referensvideoen med x264-kodaren (öppen källkod) med hjälp av ROI-informationen. ITU-standardens subjektiva test visar att den här metoden kan fungera bättre än klassisk H.264-kodning i enskilda fall.