21793 research outputs found
Sort by
What If My Body is a Beacon for the World: sensing with neurodiversity
No full textThis thesis is aimed at people who want to understand how art and collations of form feels borderless and feels twallowings* as a family of sensing and languaging. The typical thesis is the language of persons who have lots of ability to talk versus those of us who are like quiet rivers that become oceans of depth,and that is like the way I see myself so this thesis manages to vary the stream of tics and stalls (catatonia) and the man of autism is making meaning through pacing that is fast and slow and dances with the open flows of atmospheric movements.
This work shows how collaboration happens (see Appendix) and how, as an autistic non-speaking person who needs assistance to write, I need the facilitation and a poetics of the relation that goes the way of techniques and not methods. Édouard Glissant calls a poetics of relation a movement; a more-than-human relation as my body feels the atmospheres that also cue movements, impulses, ticcing, anxiety, calm. This is what I call “languaging” – an expression that exceeds words, so I have chosen a non-narrative film-making process, using tools accessible to me for a more immersive, sensory experience that I call What If My Body is a Beacon for the World: sensing with neurodiversity. What can we learn from a collaboration that not only rallies with readings, discussion, and also a “body-borrowing” that uses other hands to create? My work and life depend on collaboration and experimentation in processes that support me. Another key question: How do new grammars of neurodiverse languaging and relational pacing reconsider the spacetimes that are different from dominant perception?
*pace and movement of the stick I shake and wave making the navigations twallow are managing the way of detail
Should we stay or should we go? Clashing on Stack Exchange and Stack Overflow
No full textThe advent of Web 2.0 has opened the means of online collaboration, supporting the creation of digital platforms which, in turn, have enabled the creation and curation of knowledge commons that we sometimes take for granted. Since, researchers have revisited repertoires of contention related to digital activism, though less attention has been given to peer production and digitally native protests. This study examines the 2019–2020 crisis on the Stack Exchange network through a case study approach. Using the Governing Knowledge Commons framework, inspired by Elinor Ostrom, to structure data collection and Hirschman’s Exit, Voice, and Loyalty framework to guide analysis, the case highlights the limitations of user agency in platform governance. Despite contributors employing various strategies to voice their concerns, their ability to make themselves heard and effect change ultimately depends on administrators’ willingness to engage and their willingness to uphold platform rules, leaving contributors with limited recourse. This study argues that the absence of legal protections for knowledge commons contributors threatens the long-term sustainability of these online ecosystems. Moreover, by providing a detailed account of the crisis, this research provides data to support revising current repertoires of contention to encompass peer production contexts better
Evaluating the Use of LLMs for Automated Resolution of Web Performance Issues
No full textUsers demand fast, seamless webpage experiences, yet developers often struggle to meet these expectations within tight constraints. Performance optimization, while critical, is a time-consuming and often manual process. One of the most complex tasks in this domain is modifying the Document Object Model (DOM), which is why this study focuses on it. Recent advances in Large Language Models (LLMs) offer a promising avenue to automate this complex task, potentially transforming how developers address web performance issues.
This study evaluates the effectiveness of nine state-of-the-art LLMs for automated web performance issue resolution.
For this purpose, we first extracted the DOM trees of 15 popular webpages (e.g., Facebook), and then we used Lighthouse to retrieve their performance audit reports. Subsequently, we passed the extracted DOM trees and corresponding audits to each model for resolution.
Our study considers 7 unique audit categories, revealing that LLMs universally excel at SEO & Accessibility issues. However, their efficacy in performance-critical DOM manipulations is mixed. While high-performing models like GPT-4.1 delivered significant reductions in areas like Initial Load, Interactivity, and Network Optimization (e.g., 46.52% to 48.68% audit incidence reductions), others, such as GPT-4o-mini, notably underperformed, consistently. A further analysis of these modifications showed a predominant additive strategy and frequent positional changes, alongside regressions particularly impacting Visual Stability.
Our study highlights LLMs' clear feasibility in web performance engineering workflows, particularly for semantic concerns. However, it critically underscores the need for careful model selection, understanding their specific modification patterns, and robust human oversight to ensure reliable web performance improvements
More than words: Understanding Early Word Comprehension as a Developmental Repertoire
No full textThe ability to understand spoken words develops within the first two years of life and serves as a building block for language development. Most theoretical accounts see both maturation and language experience as essential factors in this process but differ in how they conceptualize their relationship. Some view maturation as the primary driver, others emphasize experience accumulation, while more recent frameworks suggest complex interactions between these factors. However, most accounts have been tested only through computational modelling, with limited empirical evidence examining how maturation and experience jointly predict word comprehension across different contexts. This dissertation addresses these gaps through four interconnected studies.
Chapter 2 tested competing theoretical models by investigating chronological age and language experience as predictors of performance in the looking-while-listening (LWL) paradigm. I analyzed archival data from 155 bilingual and monolingual infants, leveraging the natural variation in their language experience. An additive model—where chronological age and language experience contribute independently—emerged as the most parsimonious explanation, challenging previous accounts favouring interaction relationships.
Chapter 3 evaluated the psychometric properties of LWL outcome measures using data from 602 children. This meta-analytic investigation revealed that the paradigm captures two distinct cognitive processes—detecting word-picture mismatches versus confirming matches—suggesting word comprehension requires different response strategies depending on context.
Chapter 4 extended these findings with new data from 69 infants, examining how behavioral (looking time) and physiological (pupil dilation) responses vary across trial types. Results revealed that the age-experience relationship systematically changed based on task demands: distractor-initial trials (requiring mismatch detection) showed additive effects of maturation and experience, while target-initial trials (requiring match confirmation) were largely predicted by experience alone.
Chapter 5 investigated how family language practices shape children’s relative language exposure in bilingual households. Data from 281 bilingual children demonstrated that individual parent language use was substantially more predictive of language exposure than broader family language strategies, with mothers’ influence being twice that of fathers.
Together, these findings reconceptualize word comprehension not as a unitary skill, but as a repertoire of responses that follow distinct developmental trajectories. The framework provides a more nuanced understanding of early language development with implications for theory, assessment, and intervention
Lightweight RGB-T Object Tracking with Mobile Vision Transformers
No full textSingle-modality object tracking (e.g., RGB-only) encounters difficulties in challenging imaging conditions, such as low illumination and adverse weather conditions. To solve this, multimodal tracking (e.g., RGB-T models) aims to leverage complementary data such as thermal infrared features. While recent Vision Transformer-based multimodal trackers achieve strong performance, they are often computationally expensive due to large model sizes. In this work, we propose a novel lightweight RGB-T tracking algorithm based on Mobile Vision Transformers (MobileViT). Our tracker introduces a progressive fusion framework that jointly learns intra-modal and inter-modal interactions between the template and search regions using separable attention. This design produces effective feature representations that support more accurate target localization while achieving a small model size and fast inference speed. Compared to state-of-the-art efficient multimodal trackers, our model achieves comparable accuracy while offering significantly lower parameter counts (less than 4 million) and the fastest GPU inference speed of 122 frames per second. This thesis is the first to propose a tracker using Mobile Vision Transformers for RGB-T tracking and multimodal tracking at large
Drawing on Ecology: Artist-Teachers Using Fieldwork to Develop Arts-Based Science Journaling
No full textThis place-based master’s thesis explores how arts-based inquiry can be integrated into the elementary science education curriculum through visual fieldwork journaling. The guiding question for this research is: How can artist-teachers adapt arts-based research through visual fieldwork journals to generate place-based insights in their teaching practice? This study examines how maintaining a fieldwork journal can influence artist-teachers’ perception of place and ecological awareness within their schoolyard. It demonstrates how visual journaling can serve as a pedagogical tool for interdisciplinary learning between the arts and science.Grounded in eco-pedagogy and place-based education, this research uses a self-directed four-week fieldwork process in an urban elementary schoolyard. Using weekly prompts and structured observations, this a/r/tographic approach follows a school teacher’s exploration of four topics: mapping, living versus non-living, sounds, and water. Findings from this research indicate that combining visual and written journaling deepened ecological awareness, promoted sensory engagement, and expanded interdisciplinary connections between art and science. The four-week process revealed that drawing slowed observation, fostering an experience that supported a relational connection to the space. This research suggests that visual fieldwork journaling offers a meaningful tool for teaching and experiencing ecological literacy. It has the potential to cultivate curiosity and position students as active participants in their learning and in caring for the environment they encounter
‘A social space,’ ‘A community hub’, ‘An anchor’, ‘A refuge’: Understanding a transforming sense of place in the park through graphic journalism and (hyper)local approaches
No full textThis research-creation thesis explores the ways in which (hyper)local and graphic journalism together are able to capture a changing sense of place—the understanding of and connection to a geographic space—within a local context. It does so by applying Niblock’s “journalism practice as research”, and Chapman and Sawchuk’s “research-for-creation” and “research-from-creation”, to create a (hyper)local-graphic journalism project focused on sense of place in the Montreal neighbourhood of Notre-Dame-de-Grâce (NDG), specifically around NDG Park. This work aims to depict an evolving sense of place while reflecting on the journalist’s role in its creation.
Additionally, informed by this research’s findings and work from artist-journalists, scholarship on sense of place, and practices from first-person journalism, trauma-informed journalism and oral history, this research-creation thesis presents a set of guidelines for creating graphic journalism in a hyperlocal context
The Art of Discord: Absurdist Strategies in Contemporary Art
No full textThis thesis argues that indeterminacy, stalling, and centering absence are key aesthetic strategies for addressing the absurdity of late capitalism. These strategies do not erase, ignore, or foreclose conflict, but rather hold discord as a prerequisite to mutuality. Employing tropes from the absurdist tradition, noted for its ability to confront the spectator with apparently insoluble problems, these works do not provide didactic answers. Instead, they surface and insist on difference, disagreement, and opacity as necessary to the social fabric. In so doing, they expose the fallibility within systems of knowledge, dominant narratives, and institutions sustained through imposed consensus.
Each chapter is structured around a single strategy, through a constellation of artworks by contemporary artists from 1980–today, in the fields of performance, mixed media, and installation. Mirroring the circulation of art today, this project is inherently transnational, focusing on works by William Kentridge, Nick Cave, Rebecca Belmore, Regina José Galindo, Mierle Laderman Ukeles, Raeda Saadeh, Yinka Shonibare CBE RA, Francis Alÿs, Jill Magid, and Sophie Calle, artists working across geographies, from Canada to South Africa, in countries with their own disparate yet interrelated legacies of colonialism, resource extraction, and resistance.
Furthering Diana Taylor’s argument for the ability of performance to both record and intervene, this thesis traces the material, historical, and political conditions each singular work exists within. These strategies are critical tools for working against dominant narratives, though they are by no means inherently emancipatory. A contribution of this thesis is an increased literacy of the absurd at a time when fascist leadership globally wields its tools fluently
Do the presence of culverts and their attributes influence the spatial distribution of roadkill along Autoroute 10 and Route 112 in Southern Québec?
No full textRoads impact wildlife by fragmenting habitats, impeding movement, and though wildlife-vehicle collisions (WVCs). Proven mitigation strategies include wildlife crossing structures, but wildlife also utilize existing water culverts and underpasses to cross roadways. They offer opportunities for retrofitting when budget constraints limit the feasibility of wildlife passages. This study investigates the spatial relationship between roadkill distribution and culvert presence along a four-lane divided freeway (A10) and a two-lane regional highway (R112) in Québec, Canada. I evaluate how culvert structural characteristics (e.g., materials, shapes, functions, and dimensions) and surrounding environmental features influence roadkill frequencies separately for amphibians, mammals, and reptiles. I address three research questions: (1) Does the presence of culverts influence the spatial distribution of roadkill? (2) Which structural and environmental attributes are associated with higher or lower roadkill frequencies for different animal groups? (3) What are the similarities and differences between R112 and A10? The findings provide strong evidence for culvert influence on roadkill patterns. Mammal and reptile fatalities were more frequently associated with dry passages, amphibians with wet passages, and all groups showed a preference for using concrete structures over metal or plastic and for larger culvert openings. The results support the installation of wildlife exclusion fencing to reduce roadkill and improve driver safety. Moving forward, this study may serve as the first phase of a Before-After-Control-Impact (BACI) study, the second phase should evaluate the effectiveness of retrofitted culverts with added wildlife fencing to produce high-inference results for evidence-based WVC mitigation strategies
Proposing targets and limits to urban sprawl as an urban growth management strategy towards sustainability in Germany
No full textIn an era of rapid urbanization, achieving sustainable urban development has become a critical global concern. Yet many regions continue to suffer from the negative repercussions of urban sprawl. Unlike many other environmental issues, urban sprawl is rarely addressed quantitatively, particularly in terms of proposing concrete solutions. This study aims to bridge that gap by first measuring the extent of urban sprawl across 111 planning regions and 16 federal states in Germany using the Weighted Urban Proliferation (WUP) method for the period 1995–2015. It then introduces and analyzes five future scenarios as a framework for proposing targets, limits, and warning values for urban sprawl as a strategy to manage it. The results show that (1) Urban sprawl in Germany increased by 51.67%, rising from 2.207 UPU/m² in 1995 to 3.348 UPU/m² in 2015, 7.2 times faster than the growth in the number of inhabitants and jobs (7.2%); (2) projected population trends (whether growth or decline) have a significant influence on setting reference values and determining the sustainability of future scenarios; (3) despite projected population declines in many regions by 2050, inaction—such as continued increases in land uptake per person, as in Scenario 1—will still lead to higher urban sprawl, which highlights the importance of urban form and extent of built-up area; and (4) strategies focused on densifying built-up areas can significantly reduce urban sprawl and represent a sustainable path forward. These findings and proposed reference values provide a solid foundation for future de-sprawling efforts in regional planning and serve as a quantitative reference framework for urban planners and policymakers