Gaze tracking technology is widely used in Human Computer Interaction applications such as in interfaces for assisting people with disabilities and for driver attention monitoring. However, commercially available gaze trackers are expensive and their performance deteriorates if the user is not positioned in front of the camera and facing it. Also, head motion or being far from the device degrades their accuracy. This thesis focuses on the development of real-time time appearance based gaze tracking algorithms using low cost devices, such as a webcam or Kinect. The proposed algorithms are developed by considering accuracy, robustness to head pose variation and the ability to generalise to different persons. In order to deal with head pose variation, we propose to estimate the head pose and then compensate for the appearance change and the bias to a gaze estimator that it introduces. Head pose is estimated by a novel method that utilizes tensor-based regressors at the leaf nodes of a random forest. For a baseline gaze estimator we use an SVM-based appearance-based regressor. For compensating the appearance variation introduced by the head pose, we use a geometric model, and for compensating for the bias we use a regression function that has been trained on a training set. Our methods are evaluated on publicly available datasets.

Many construction companies today utilize webcams on their jobsites to monitor and record construction operations. Jobsite monitoring is often limited to outdoor construction operations due to lack of mobility of wired webcams. A wireless webcam may help monitor indoor construction operations with enhanced mobility. The transfer time of sending a photograph from the wireless webcam, however, is slower than that of a wired webcam. It is expected that professionals may have to analyze indoor construction operations with longer interval time-lapse photographs if they want to use a wireless webcam. This research aimed to determine the maximum time interval for time-lapse photos that enables professionals to interpret construction operations and productivity. In order to accomplish the research goal, brickwork of five different construction sites was videotaped. Various interval time-lapse photographs were generated from each video. Worker?s activity in these photographs was examined and graded. The grades in one-second interval photographs were compared with the grades of the same in longer time interval photographs. Error rates in observing longer time-lapse photographs were then obtained and analyzed to find the maximum time interval of time-lapse photography for monitoring construction operations. Research has discovered that the observation error rate increased rapidly until the 60-second interval and its increasing ratio remained constant. This finding can be used to predict a reasonable amount of error rate when observing time-lapse photographs less than 60-second interval. The observation error rate with longer than 60-second interval did not show a constant trend. Thus, the 60-second interval could be considered as the maximum time interval for professionals to interpret construction operations and productivity.

Untersucht wurde, ob und wie Webcams auf privaten Homepages im Internet als Selbstdarstellungs-Medium genutzt werden und welchen Einfluss dabei Persönlichkeitsmerkmale und Ziele von Webcam-Nutzern ausüben. An einer internetbasierten Befragung nahmen 62 Personen teil, die über eine zufallsgesteuerte Stichprobe ermittelt wurden. In der Befragung wurden die Persönlichkeitsvariablen der Big Five und Authentizität erhoben, sowie interessierende Facetten des Narzissmus, der Selbstwertschätzung, der Selbstüberwachung sowie neun Stile der Selbstdarstellung. Darüber hinaus machten die Teilnehmer Angaben zu Webcam-Motiven, Gründe für die Installation der Webcam, Zielgruppen und Nutzungsweisen von Webcams sowie zu weiteren interessierenden webcam- und homepage-spezifischen Merkmalen. Erste Ergebnisse zeigen, das Webcams auf privaten Homepages zur authentischen Selbstdarstellung verwendet werden. Weiterhin unterscheiden sich Webcam-Besitzer nur marginal hinsichtlich der erhobenen Persönlichkeitsvariablen von Vergleichsstichproben und von Homepage-Besitzern. Dabei richten sich Webcam-Besitzer vorrangig an ein persönlich bekanntes oder nichtbekanntes Publikum und verwenden auch im realen Leben bevorzugte Selbstdarstellungsstile. Abschließend wird die Darstellung der Webcam-Nutzung in öffentlichen Medien kritisch disskutiert und Anregungen für die weitere Forschung in diesem Bereich gegeben.

[pt] As comunicações à distância estão crescendo consideravelmente pelo uso de aplicações através da Internet e de ambientes virtuais. Interaçõoes sociais e pessoais têm recebido especial enfoque, sobretudo a videoconferência, acarretando uma grande demanda de tecnologia apropriada para esses sistemas. Resultados de pesquisas em Presença, alguns deles obtidos nesta tese, permitiram inferir os quatro pilares da proposta: 1. visualização é muito importante nas interações sociais/pessoais; 2. tecnologia simples pode promover presença social; 3. imagens de câmera são mais satisfatórias do que avatares para algumas interações sociais/pessoais; 4. estereoscopia influência presença positivamente. Sob a restrição de usar tecnologia simples e de baixo custo, a tese propõe o uso de visualização estereoscópica para webcams, em tempo real. Porque webcams são equipamentos simples com baixa resolução de imagem, a obtenção de um estéreo com qualidade torna- se um desafio. Partindo do princípio que o olho humano não se comporta da mesma forma para ver de perto e de longe, a pesquisa analisou o processamento de estéreo natural do olho humano, e com base em resultados psicofísicos e fisiológicos da visão binocular, a tese propõe um modelo para o processamento de imagens estereoscópicas a serem visualizadas em curta distância, e usa este modelo na implementação apresentada. Os resultados obtidos foram satisfatórios para a visualização de imagens estéreo com webcams, em tempo real, e principalmente, eliminaram a necessidade de reajuste de paralaxe das imagens e do reposicionamento das câmeras a cada movimentação do observador, dando assim mais liberdade de visualização ao observador. / [en] The improvement of communication among remotely located people constitutes a very comprehensive prominent subject nowadays. Among applications to attain this goal, some aim to support social and personal interactions to meet partners, friends and family in distance. Presence research results, some obtained as part of this thesis, allowed to infer the four pillars of the proposal: 1. visualization is very important in social/personal interactions, mainly when people have close social/personal relationship; 2. simple technology can afford social presence; 3. camera images are more satisfactory than avatars during some social/personal interactions; 4. stereoscopy influences presence positively. Under the constraint of using simple technology without adding any special hardware to a PC system, the thesis proposes the use of stereoscopy through webcams, in real- time, as one solution to add 3D features to camera images. Since webcams are simple equipments with low resolution, to obtain stereo images in good quality is a challenge. Considering the fact that human eyes treat short and long distance in a different way, this research analysed the natural stereo processing of human eye, and based on psychophysical and physiological features of binocular vision, the thesis proposes a model to process stereoscopic images to be visualized in short distance, and implemented it as a proof of concept. The obtained results are satisfactory for stereoscopic images visualization through webcams, in real-time, and mainly, eliminated the need of adjusting images parallax and cameras position for each viewer´s movement, allowing more freedom to the viewer.

This study was developed to investigate the readiness or willingness of providers, speech language therapists (SLPs), and patients to participate in telepractice for speech-language therapy services in rural areas of Nevada. Nevada has not ranked high in healthcare for several years and was noted to have low SLP to patient ratios, especially in rural areas of Nevada. A thorough literature review identified international and national studies that have shown resistance to telepractice expressed by patients and providers, but it has also shown an increase in interest post telepractice therapy. Nevada does not have current research on readiness to participate in telepractice medical settings of rural areas of Nevada, therefore a survey was developed to ask providers and patients for their opinions about participating in telepractice. Providers were selected using an online directory search, and patient participants were chosen at the discretion of the providers. A unique survey was presented to each type of participant (provider or patient) in order to compare opinions and readiness of each participant type. Specific questions related to voice and swallowing disorders were addressed. A large sample size was not obtained, but the researcher was able to make some associations to readiness to participate in telepractice and possible factors that may or may not have affected a participant’s opinion. Age of participants, type of internet used, availability of webcam, and area of the state did not appear to have an association with the readiness to participate in telepractice. A positive response rate to engaging in telepractice of 40% was noted across patients and providers. Though the participant sample size was small, there is still some indication that telepractice could be used to address the needs of patients who need speech-language therapy services in rural areas of Nevada.

This thesis deals with the possibility of using webcams as a source of spatial data for snow occurrence. The aim of this study is to propose a suitable method of snow cover detection from web camera images. From a sample of 6 webcams of the Czech Hydrometeorological Institute (CHMI) the snow cover is detected by pixel classification methods. The effect of training file size on the accuracy of classification is examined and the overall accuracy achieved by the SVM method was shown to be 97.46%. This study also aims to propose a system for determining the proportion of snow-covered areas. The algorithm consists of several sub-steps: filtering and registration of images, detection of snow, introduction of a coordinate system, calculation of the size of the surveyed area and the proportion of snow-covered area. The designed model can be used for automatic processing of images for various webcams. The melting curves of the snow cover are generated from the obtained daily values of the snow covered area. The results are validated using data from selected CHMI stations. The proposed and parameterized model confirms the possibility of successful use of webcams as a complement to ground measurement of meteorological stations and for the validation of remote sensing products.

Zugl.: Zürich, Techn. Hochsch., Diss., 2009

This dissertation explores the impact of technology on social life. Focusing on webcam-mediated audio-visual conversations between Japanese families in the United States and their extended family members in Japan, I examine how technology participates in creating an interactional space for the families to manage intra- and intergenerational relationships. Combining ethnography with turn-by-turn analyses of naturally occurring webcam interactions, I specifically investigate how cultural, discursive, and family practices are transformed in innovative ways and how families adapt to the emerging mediated space. Looking at how interactional activities are coordinated across spaces, I show that webcam interactions constitute a new type of shared living space in which multigenerational family relationships are created and managed. I call this emerging space the virtual ie (‘house’ and ‘stem-family system’). In this virtually conjoined space, children, parents, and grandparents are visually familiarized with each other’s households and socialized to each other’s virtual presence. Even though the ie is no longer a juridical unit of co-residence, my goal is to discuss the significance of the ie in understanding how transnational Japanese families can dwell in a shared living space created by webcam interactions. My analyses demonstrate how webcam encounters create a stage for participants to perform various identities in interactions. Learning to talk and participate in such webcam interactions, children are socialized to their ie belongings and identities. Additionally, even though far-flung children do not provide physical and daily care for their parents in Japan, they actively take care of elder parents’ media environments. I demonstrate that what I call media care practices add another context for adult children living abroad to carry out their filial responsibilities. I also show that the management of webcam visual fields creates a type of social field that reflects local understandings of social positioning in ie structures. How participants decide to display themselves to others by manipulating the webcam’s visual fields provides a new way to demonstrate various social relationships and responsibilities over long distances. From this perspective, a virtual ie is not merely a reflection of an ideological understanding of Japanese families, but an interactional achievement facilitated by webcams. / text

Deep understanding of urban mobility is of great significance for many real-world applications, such as urban traffic management and autonomous driving. This thesis develops deep learning methodologies to extract vehicle counts from streaming realtime video captured by multiple low resolution web cameras and construct maps of traffic density in a city environment; in particular, we focus on cameras installed in the Manhattan borough of NYC. The large-scale videos from these web cameras have low spatial and temporal resolution, high occlusion, large perspective, and variable environment conditions, making most existing methods to lose their efficacy. To overcome these challenges, the thesis develops several techniques: 1. a block-level regression model with a rank constraint to map the dense image feature into vehicle densities; 2. a deep multi-task learning framework based on fully convolutional neural networks to jointly learn vehicle density and vehicle count; 3. deep spatio-temporal networks for vehicle counting to incorporate temporal information of the traffic flow; and 4. multi-source domain adaptation mechanisms with adversarial learning to adapt the deep counting model to multiple cameras. To train and validate the proposed system, we have collected a largescale webcam traffic dataset CityCam that contains 60 million frames from 212 webcams installed in key intersections of NYC. Of there, 60; 000 frames have been annotated with rich information, leading to about 900; 000 annotated objects. To the best of our knowledge, it is the first and largest webcam traffic dataset with such large number of elaborate annotations. The proposed methods are integrated into the CityScapeEye system that has been extensively evaluated and compared to existing techniques on different counting tasks and datasets, with experimental results demonstrating the effectiveness and robustness of CityScapeEye.

Movement-based games are becoming more frequent in everyday lives. With easier access to webcams built into laptops and web-based games, a game which utilises both concepts can become a good option for everyday gaming.This study evaluates a movement-based game written in JavaScript using phaser.io which uses a webcam for control and HTML5 technique for capture. The main questions in the study are what observations can be done and are there any problems when developing a game with webcam-based motion-detection in regards to factors like flow, feedback and positioning. This study tries to answer these questions by building upon an existing game and implementing new functions and feedback and then observing how these effect the game-play. The results of the study showed that it is important to consider good feedback and how to position oneself when playing such a game and that more testing should be done to gain further knowledge about these two. But also that many of the implementations done made the person achieve flow while playing. Further work should be beneficial also to make the game even better while keeping it easy to play.