Theses, Dissertations, and other Required Graduate Degree Essays

Activity analysis in which multiple people interact across a large space is challenging due to the interplay of individual actions and collective group dynamics. We propose an end-to-end generic approach for learning person trajectory representations for group activity analysis. The learned representations encode rich spatio-temporal dependencies and capture useful motion patterns for recognizing individual events, as well as characteristic group dynamics that can be used to identify groups from their trajectories alone. We develop our deep learning approach in the context of team sports, which provide well-defined sets of events (e.g. pass, shot) and groups of people (teams). We evaluate our model on NBA basketball and NHL hockey games datasets. Analysis of events and team formations using these two sports datasets demonstrate the generality of our approach. Experiments show that our model is capable of (1) capturing strong spatio-temporal cues for recognizing events in hockey dataset (2) capturing distinctive group dynamics for identifying group identity.

The use of Bayesian statistical methods to handle missing data in biomedical studies has become popular in recent years. In this thesis, we propose a novel Bayesian sensitivity analysis (BSA) model that accounts for the influences of missing outcome data on the estimation of treatment effects in randomized control trials with non-ignorable missing data. We implement the method using the probabilistic programming language Stan, and apply it to data from the Vancouver At Home (VAH) Study, which is a randomized control trial that provided housing to homeless people with mental illness. We compare the results of BSA to those from an existing Bayesian longitudinal model that ignores missingness in the outcome. Furthermore, we demonstrate in a simulation study that, when a diffuse conservative prior that describes a range of assumptions about the bias effect is used, BSA credible intervals have greater length and higher coverage rate of the target parameters than existing methods, and that sensitivity increases as the percentage of missingness increases.

Activity monitoring is the task of continual observation of a stream of events which necessitates the immediate detection of anomalies based on a short window of data. For many types of categorical data, such as zip codes and phone numbers, thousands of unique attribute values lead to a sparse frequency vector. This vector is then unlikely to be similar to the frequency vector obtained from the training set collected from a longer period of time. In this work, using topic models, we present a method for dimensionality reduction which can detect anomalous windows of categorical data with a low rate of false positives. We apply nonparametric Bayesian topic models to address the variable nature of data, which allows for updating the model parameters during the continual observation to capture gradual changes of the user behavior. Our experiments on several real-life datasets show that our proposed model outperforms state-of-the-art methods for activity monitoring in categorical data with large domains of attribute values.

Three dimensional (3D) videos are the next natural step in the evolution of digital media technologies. In order to provide viewers with depth perception and immersive experience, 3D video streams contain one or more views and additional information describing the scene's geometry. This greatly increases the bandwidth requirements for 3D video transport. In this thesis, we address the challenges associated with delivering high quality 3D video content to heterogeneous devices over both wired and wireless networks. We focus on three problems: energy-efficient multicast of 3D videos over 4/5G networks, quality-aware HTTP adaptive streaming of free-viewpoint videos, and achieving quality-of-experience (QoE) fairness in free-viewpoint video streaming in mobile networks. In the first problem, multiple 3D videos represented in the two-view-plus-depth format and scalably coded into several substreams are multicast over a broadband wireless network. We show that optimally selecting the substreams to transmit for the multicast sessions is an NP-complete problem and present a polynomial time approximation algorithm to solve it. To maximize the power savings of mobile receivers, we extend the algorithm to efficiently schedule the transmission of the chosen substreams from each video. In the second problem, we present a free-viewpoint video streaming architecture based on state-of-the-art HTTP adaptive streaming protocols. We propose a rate adaptation method for streaming clients based on virtual view quality models, which relate the quality of synthesized views to the qualities of the reference views, to optimize the user's quality-of-experience. We implement the proposed adaptation method in a streaming client and assess its performance. Finally, in the third problem, we propose an efficient radio resource allocation algorithm in mobile wireless networks where multiple free-viewpoint video streaming clients compete for the limited resources. The resulting allocation achieves QoE fairness across the streaming sessions and it reduces quality fluctuations.

Aspergillus fumigatus is a filamentous fungus that is the most common cause of life-threatening invasive mould infections in immunosuppressed individuals. A. fumigatus produces a sialidase enzyme that shows a preference for 2-keto-3-deoxy-D-glycero-D-galacto-nononic acid, (KDN). Sialidases break the glycosidic bond between terminal sialic acids and an underlying glycan chain. The purpose of my research was to create and characterize the KDNase knockout and complemented strains in A. fumigatus. Both strains were successfully generated. Growth in the presence of cell wall stressors (hyperosmolarity, antifungal agents, Congo Red dye) showed that the KDNase gene deletion affected morphology and cell wall integrity. Treatment of A. fumigatus conidia with endogenous AfKDNase enzyme resulted in conidial clumping and damage, an effect not observed when conidia were treated with a bacterial sialidase. The Δkdnase strain remained virulent in an immunosuppressed murine model of invasive aspergillosis.

Anthropocentric views historically have limited the potential of respect for other-than-human beings by promoting ideologies of human exceptionalism with regard to consciousness, reason and language. The doctrine of human supremacy has become normalized in capitalistic cultures, driving the domination and exploitation of other beings and natural systems for their ‘use-value’ as ‘resource,’ leading to today’s catastrophic harms of climate change, species extinction, ecological degradation, ocean acidification, industrial farming, and animal slavery. As a means to counteract anthropocentrisms, this thesis proposes biophilic ethics and its constituent details—communicative ethics, biophilic attention, intentional relationality, interspecies generative indeterminacy—explored through art-action. The interdisciplinary investigation looks at methodologies in philosophy, ethics of care, ecofeminism, cognitive ethology, biology, naturalist methods, and aesthetics that interrogate beliefs in human superiority, and propose relational approaches to situate the human alongside Earth’s other beings within our shared ecosystems. The epistemological investigation is woven into ontological explorations rooted in relational events that happened while conducting interspecies processes in my art practice over the past decade. Each creative instance with dogs, crows and stones, fishes, and forests is examined for potential towards ecological understanding and compassionate action. Four projects that emerged from the interspecies processes—EPIC_Tom, Crow Stone Tone Poem, Salmon People, and Biophilia—model thinking-feeling and responding to Earth’s more-than-human co-inhabitants. The thesis expands on previous thought with regard to biophilic ethics by arguing that love for life is a lived-condition beyond human-centred values. Other beings are themselves biophilic—interested in their lives, their families, communities, and cultures. This expansion of biophilic ethics is explored through potentials emerging in more-than-human relational encounters enacted using ethics of care methods. Ethics of care calls into question objectivist and utilitarian methods, and instead promotes empathy practices to sense and feel our interdependent experiences. The attendant details of biophilic ethics proposed in the thesis are adaptations of previous thought, expanded through historical and contemporary aesthetic methods. Communication ethics calls for improved relations with other animals through developments in understanding about differing modes of perception and communication. My expansion on this thought is explored through applied interactions with other beings for interspecies expression and communication potentials. Biophilic attention is a development of aesthetic observational techniques, and proposes expanded attention towards other beings in relation to one’s own sensing-feeling-thinking in the world. Intentional relationality is proposed as respect for differing minds and life projects, emerging through interspecies participatory art methods that provide open-ended collaborative inquiry. Interspecies generative indeterminacy is informed by thought on agential intra-action combined with generative and indeterminacy methods first articulated in mid-twentieth century art practices and more recently explored in computational art. These details of biophilic ethics are examined within ranges of locatedness—the home, the neighbourhood, the territory, the Earth.

The brain is perhaps one of the most fundamental organs in all vertebrates. It determines not only an individual’s ability to sense and process stimuli from the environment, but is also crucial in maintaining internal homeostatic processes as well as determining an individual’s cognitive abilities. Brains come at a steep energetic cost however, with neural tissue requiring ~20 times the energy of muscle tissue. With such an important role to play, the ‘expensive brain hypothesis’ was been established to understand the evolutionary correlates of brain size. Maternal investment, defined as energetic investment during development, is a strong underlying factor in brain size evolution where higher energy investment from mothers is associated with increased brain size. However, much of what we know about brains comes from studying birds and mammals, while generally overlooking other vertebrate classes. Despite their diversity, all jawed vertebrate brains are comprised of similar components, a pattern that first appeared in sharks, rays, and chimaeras (Chondrichthyans). Chondrichthyans are often disregarded as unremarkable from a comparative perspective, which overlooks their true diversity of life histories and ecological niches. This thesis seeks to understand the evolution of brain size and organization in relation to life history and maternal investment using chondrichthyans as a model system. First, I reveal the sequence of reproductive evolution, finding that egg-laying is ancestral and that live-bearing and additional maternal investment (matrotrophy) have evolved independently several times, and are correlated with increasing body size. Second, I find that the evolution of reproductive mode and ecological lifestyle underlie the evolution of both brain size and brain organization, such that shallowwater matrotrophic species have large brains that are predominantly composed of regions related to enhanced cognitive abilities, the telencephalon and cerebellum. Conversely, deepwater lecithotrophic species have small brains composed predominantly of medulla oblongata. Lastly, I find that similar patterns of regional scaling in mammals, birds and chondrichthyans differ from those of teleosts, agnathans, and amphibians, and I propose that differing reproductive strategies may underlie this variation.

In Afghanistan, a significant advance since the fall of the Taliban has been the entry of women into media as reporters, anchors and producers. Media, in essence, have become a battleground for Afghan women fighting to overcome a culture of silence and invisibility following years of oppression. Be it TV, newspapers, radio or even music, media allow for the dissemination of stories that speak to women’s social, economic and political realities. Mightier Than The Sword is a two-part project exploring this social advancement. The first component is a 47-minute documentary, shot in Afghanistan in 2015, analyzing how the work of female journalists has affected gender perceptions and gender equality. The written portion of Mightier Than The Sword is an in-depth examination of the history of the media in Afghanistan and the effect mass media have had on gender perceptions in Afghanistan. First-hand interviews, conducted in 2015, are included.

Field mapping, petrography, thermodynamic modelling, and U-Pb (monazite and zircon) and 40Ar/39Ar geochronology reveal the tectonometamorphic history of polydeformed, amphibolite-facies rocks near Aishihik Lake, Yukon. Detrital zircon U-Pb ages suggest that these rocks are correlative to the Snowcap assemblage of the Yukon Tanana terrane. A penetrative regional foliation (S1) developed during the late Paleozoic, as S1 is cross-cut by a late Permian pluton. Permian plutons also exhibit less strain than Mississippian plutons near Aishihik Lake. The main foliation (S2) reflects west-verging, ductile shear (D2) during amphibolite facies metamorphism. Dating of Low-Y metamorphic monazite constrains the timing of D2 to 200-190 Ma. Peak T and P during D2 were 640-650 °C and ~7 kbar, respectively. High-Y monazite ages date regional decompression at ca. 188 Ma. 40Ar/39Ar thermochronology results indicate regional cooling through muscovite closure at ca. 175 Ma, whereas ca. 126 Ma biotite may reflect cooling following east-verging Jura-Cretaceous deformation (D3).

Systems of resort governance do not emerge in a vacuum, instead they are the product of forces and the will of individuals. This study examines the emergence and evolution of resort governance systems. Using British Columbia as a case study, the research explores the driving forces which influenced the creation of significant provincial policies and pieces of legislation that comprise the regional resort governance system. Critical moments in the evolution of British Columbia’s resort governance are explored to betterunderstand the impact of these forces and how they were negotiated. Employing a path creation lens, the project illustrates the importance of past decisions and the power of strategically leveraging forces.