21793 research outputs found
Sort by
On Addressing Psychological Intimate Partner Violence Against Women Through a Change of Educational Curricula and Restorative Practice Suggestions
Get PDFThis master's thesis explores the educational and social approaches to addressing psychological intimate partner violence (IPV) against women in Québec. By examining current educational curricula and integrating restorative practices suggestions, the study highlights how changes in educational frameworks can contribute to reducing IPV in society (especially psychological IPV). Through a combination of a literature review and qualitative data analysis derived from interviews with IPV victims and social workers, this research identifies key strategies for promoting awareness about and prevention of psychological IPV. The findings emphasize the importance of integrating these practices into educational systems to foster a more supportive and responsive approach to psychological IPV
Adaptive Correction Strategy in Robotic Gas Tungsten Arc Welding for Additive Manufacturing
Get PDFWire arc additive manufacturing (WAAM) is one of additive manufacturing (AM) methods and owns notable advantages like enabling production of large-scale components. However, the current WAAM has inherent drawbacks, such as heat accumulation and near-net-shape production issues that can lead to defects like geometrical deviations, porosity inside the weld and/or surface irregularities. It is noted that various process parameters are directly related to the above-mentioned production issues. Therefore, it is crucial to set various process parameters (based on geometry, processing changes when depositing, etc.) appropriately for achieving a good quality of product
In this project, we aimed to investigate the influence of various process parameters on the product quality and to automate the WAAM process using a vision system. To realize the objectives, we have developed an adaptive correction strategy to control the robotic welding machine, i.e. a Nertamatic power source that centralize the welding cycle while considering various welding parameters such as the robot path, deposited layer height, surface contamination, etc. To minimize operator intervention, a Cognex 3D A5000 series camera was employed to scan/monitor the deposited object layer by layer. The camera’s ASCII output was used for mesh processing. An on-line control scheme has been developed to control the robot path according to the dimensions of previous layers and thus the robot’s height was adaptively adjusted. An algorithm was proposed to process the mesh data to detect and correct the inconsistencies by commanding the robot to stop in case of collisions or fill cavities in the case of insufficient height or underfills.
An experiment has been designed on a robotic WAAM machine where a TopTig gun was attached to the end effector of a 6-degree-of-freedom (DOF) ABB robot IRB4600, equipped with a 2-DOF IRBP_A500 table, to deposit material layer by layer. A nozzle is mounted with different diameter tungsten electrodes and fed with various wire materials. In this study, we tested a 3 mm diameter tungsten electrode and stainless-steel filler wire to assess the effects of overlap between the beads on the integrity of the deposited part. Simulations have been conducted in RobotStudio software to validate the recognizing deposited layer inconsistencies and the effectiveness of the path planning and welding machine settings, demonstrating the potential for deploying the adaptive correction strategy in the WAAM process
Oil Spill detection and Fingerprinting Using Semantic Segmentation and Data-driven Modeling
Get PDFOil spills significantly threaten marine environment, damaging ecosystem, wildlife, and coastal communities. This thesis addresses these challenges by employing both advanced machine learning techniques and satellite imagery analysis technologies to enhance the accuracy and efficiency of oil spill detection and or source identification. By utilizing Synthetic Aperture Radar (SAR) images and examining semantic segmentation models, the research aims to accurately detect oil spills based on satellite images. Additionally, oil fingerprinting techniques, involving unsupervised classification are used to identify the sources of marine oil spills, providing a comprehensive framework for oil spill monitoring and management. The methodology involves the use of three distinct datasets: a multi-class dataset for detecting oil spills using satellite images, a binary dataset focusing on oil spill incidents in the Gulf of Suez from 2017 to 2021 as a case study for oil spill detection and a dataset for oil fingerprinting based on samples from the MV Manolis L shipwreck. For oil spill detection, semantic segmentation models were trained and evaluated using these datasets. Performance metrics such as Intersection over Union (IoU) were used to assess the modeling accuracy. Secondly for oil fingerprinting, PCA and HCA were applied to analyze the chemical composition data of the MV Manolis L. oil samples to identify their similarities and differences for oil source classification.
The results indicate that DeepLabv3+ and UNet++ models achieved the highest mean Intersection over Union (mIoU) scores for multi-class and binary segmentation tasks, respectively, demonstrating their robustness in detecting oil spills. Specifically, DeepLabv3+ achieved a mIoU of 68.3% in the multi-class dataset, excelling in complex categories like oil spills and look-alikes. UNet++ achieved a mIoU of 87.5% in the binary dataset, highlighting its effectiveness in distinguishing oil from non-oil regions. For oil fingerprinting, the Support Vector Classifier (SVC) model exhibited the highest accuracy, particularly in predicting the composition of n-alkanes, PAHs, and TPH, with F-scores of 1.0, 0.987, and 0.975, respectively. These findings underscore the effectiveness of coupling advanced machine learning models with established chemical analysis techniques, offering a reliable approach for oil spill detection and the subsequent effective cleanup
Improving the Program Revision Process Using CourseFlow, a Digital Educational Tool
Get PDFThis research examines the impact of CourseFlow, a cloud-based educational tool, in the program revision process at the Physiotherapy Technology program at Dawson College in Montreal, Canada. Through semi-structured interviews with key stakeholders involved in program revision, the research looks at how CourseFlow supports practitioners in improving their own professional practices, fostering collaboration, and aligning course content with program outcomes, overall improving curriculum mapping and program revision processes. The research also explores how CourseFlow compares to other tools commonly used in program revisions, highlighting both its advantages and disadvantages. Some of the findings highlight that while CourseFlow enhances communication, facilitates more meaningful discussions, and promotes reflective practices within the program revision process, at the same time, it has limited editing capabilities, experiences difficulties with content migration, and provides for a steep learning curve for new users. Recommendations for improving CourseFlow include refining its visual interface, enhancing collaboration features, and enabling more effective side-by-side course comparisons. These improvements aim to further optimize CourseFlow for broader adoption and use. Suggestions for future studies outlined in this research include revisiting the original research design post-pandemic and utilizing the unused data collected during this study
Effects of sound masking noise on workers’ perception and performance
Get PDFThere is always a lack of separation between individual work areas in open offices, so sound insulation is poor. A cheap and effective solution to this problem may be to use sound masking technology. This study aims to explore the impact of sound masking noise on employee perception and psychology in scenarios involving two types of signal speech noise involving sound masking technology. Specifically, this study implemented background noise, speech, visual stimuli, and interactive components required in a real-time virtual reality framework. Ten participants participated in a multi-task cognitive experiment. The key metrics for evaluation include task completion rate, accuracy, NASA Task Load Index, and individual noise sensitivity scores.
The average accuracy is lower in the 50 dBA with speech condition compared to the 38 dBA with speech condition. Similarly, participants complete fewer math questions on average in the 50 dBA with speech condition compared to the 38 dBA with speech condition. The combination of higher noise levels and speech (bad signal-to-noise ratio for speech) significantly hampers task efficiency. It shows that the ten participants have different levels of sensitivity to noise or speech. People with higher noise sensitivity will experience the highest task load in noisy environments with speech. In environments with irregular noise patterns, sound masking systems may inadvertently amplify rather than mask irregular noise. This study not only observed individual differences in noise sensitivity and cognitive effects, but also highlighted the importance of individual responses to noise in managing noise exposure in the open plan offices.
Keywords: sound maskin
In/Convenience: Inhabiting the Logistical Surround
Get PDFConvenience is the feeling and aspiration that animates our platformed present. As such, it poses urgent techno-political questions about the everyday digital habitus. From next-day delivery, gig work, and tele-health to cashless payment systems, data centers, and policing – convenience is an affordance and an enclosure; our logistical surround. Driving every experience of convenience is the precarious work, proprietary algorithms, or predatory schemes that subtend it. This collaborative book traces how the logistical surround is transformed by thickening digital economies and networked rituals, examining contemporary conveniences across a wide range of practices and geographies. Contributors examine the ineluctable relation between convenience and its constitutive opposite, inconvenience, considering its infrastructural, affective, and compulsory dimensions. Living in convenience is thus both a hyper visible manifestation of so-called late capitalism and a pervasive mood that fades into the background (like the data centers that power it). Bringing the agonistic relation of in/convenience to center stage, this volume analyzes the logistics of delivery, streaming porn, cloud computing, water infrastructures, smartness paradigms, convenience stores, sleep apps, surveillance, AI ethics, and much more – rethinking the cultural politics of convenience for the present conjuncture
Behavioural and Neural Analyses of Higher-Order Fear Conditioning
Get PDFMemories about aversive events that elicit fear can imbue fear to other stimuli. This is studied using Pavlovian higher-order fear conditioning. A stimulus that is directly paired with an aversive outcome (i.e., Pavlovian first-order fear conditioning) can support learning about another stimulus (i.e., Pavlovian higher-order fear conditioning). That is, by virtue of its links with a first- order stimulus (i.e., S1), a higher-order stimulus (i.e., S2) controls behaviour by eliciting conditioned defensive responses. This occurs in two ways: either S2 is paired with S1 before the latter is paired with an outcome (i.e., foot shock) as exemplified in sensory preconditioning (SPC), or after S1 is paired with an outcome as exemplified in second-order conditioning (SOC). Reduction in fear to S1 by presenting it in the absence of the shock transfers to sensory preconditioned but not to second-order conditioned fear, showing that two types of fear are supported by distinct behavioural (and neural) structures.
Extensive work investigated the neural substrates underlying first-order fear conditioning, while our understanding of the neural structures mediating higher-order fear conditioning is still relatively limited. The present thesis investigated the role of some of the neural substrates that support sensory preconditioning and second-order conditioning. In Chapter 4 we confirmed that SPC but not SOC required the integrity of first-order fear and provided neural evidence for this dissociation using a chemogenetic approach to delete first-order fear memory, which disrupted SPC but not SOC. In Chapter 5 we investigated the role of lOFC in regulating the expression of both types of fear and showed that lOFC inactivation prior to test disrupted SPC but enhanced SOC. In Chapter 6 we identified the neuronal ensembles that are activated by SPC and SOC in the BLA, a region critical for the expression of fear to SPC and SOC, and showed that a subset of these ensembles showed projections to the lOFC. These projections, when silenced disrupted both types of fear. Lastly, we silenced lOFC input to the BLA and showed that this pathway is crucial for SPC but not SOC. Our findings delineate neurobiological structures differentially supporting SPC and SOC types of fear and characterize the role of lOFC in the fear circuit
Stuck on the Wrong Side of the Tracks: Crime and Neighbourhood Change Across Adulthood
Get PDFMoving from a disadvantaged neighbourhood to one of more affluence has been shown to improve life outcomes. However, not everyone manages to overcome the environmental and social hazards of such neighbourhoods. Success may depend on individual differences such as childhood social behaviour, education, and criminal activity. Crime and neighbourhood disadvantage are highly correlated, but the directional nature of this relationship and its transactional nature throughout life have rarely been examined. Part One of the current investigation examined whether individual characteristics, including childhood social behaviour, education, and criminality, contribute to the perpetuation of socioeconomic immobility across adulthood via neighbourhood disadvantage using a growth curve model. In Part Two, the potential transactional nature of associations between crime and disadvantage over time were examined utilizing a cross-lagged analysis.
Participants were drawn from the Concordia Longitudinal Research Project, a prospective, 47-year longitudinal investigation of over 4000 families from neighbourhoods of low socioeconomic status in Québec, Canada. In Part One, Growth curves modeled differences in change in participants’ neighbourhood disadvantage (via census data) over 30 years, from middle-childhood (age 7-12) to middle-adulthood (age 46-57). Predictors included childhood social behaviours and total criminal charges in early adulthood (age 18-28). In Part Two, to examine potential transactions, cross-lagged associations were modeled between neighbourhood disadvantage across four time points (1976, 1986, 1996, 2006). In this model, childhood neighbourhood disadvantage (1976) and aggression were included as predictors and total years of education was included as a mediator.
Part One results indicated that participants with no criminal charges showed the greatest improvement in neighbourhood over time, whereas those with many charges showed little improvement. Participants with histories of childhood aggression, withdrawal, or lower likeability were also less likely to experience improvements. Results from Part Two indicated that the association between charges and neighbourhood disadvantage was transactional over time and that education may play an important protective role for individuals who grow up in disadvantaged neighbourhoods or for more aggressive children. These findings provide evidence for the importance of criminality in undermining at-risk young adults’ ability to overcome neighbourhood disadvantage, highlighting risk and protective factors that may inform early and long-term intervention and policy
Yarn's Not Dead and Neither Are You: A Punk Knitter's Journey to Uncovering the Soft and Squishy Superpowers of Knitting
Get PDFIn 2022, I moved to Paris and everything in my life fell apart. Little did I know, knitting was going to become my lifeline. This is a research-creation thesis that describes my journey through a non-linear, unconventional healing process that led me to realize that there are so many benefits to being a knitter, on both the personal and educational aspects of life. The research is supported by a theoretical foundation of punk scholarship. I argue that punks share a lot with knitters in many aspects of their ethos. Most prominently, they both consider DIY as their core value. Furthermore, I believe that they both display subversive behaviors by not conforming to the established ideas of commercial beauty. The creative portion of this project is made up of 12 shawls I have knitted in the past two years. These woolly, emotionally complex landscapes of fears and self-doubt, but also of joyful reminiscence and precious memories, have provided me with an alternative way to process inner darkness by engaging with a colorful, comforting material, as well as highlight the fact that knitting has a lot of pedagogical value, in both formal and informal schooling structures. Together, the shawls form a soft, chaotic and powerful reminder that (a), healing takes many forms, (b), there are alternative ways to develop computational skills and (c), it is ok to be punk, academic and crafty all at once
Bytecode Similarity Detection for Obfuscated Java Android Applications
Get PDFCode similarity detection has many practical applications, such as intellectual property protection, vulnerability search, and malware detection. However, existing approaches typically focus on the source code, while many third-party libraries are released in bytecode format. Hence, developers may unknowingly use third-party libraries without knowing possible license violations or vulnerabilities. In this thesis, we introduce a deep learning approach, ByClone, to detect source code clones based on Java bytecode. We collect source-code level clone data for bytecode in 140 Android applications to conduct the experiments. We find that ByClone is effective in detecting code clones based on bytecode, with a precision and recall of 78.37 and 75.24. After obfuscating the bytecode, ByClone still has a precision and recall of 82.55 and 70.95, highlighting the potential of ByClone. Finally, we find that ByClone is not sensitive to different obfuscation options. Our study highlights the potential of clone detection based on bytecode. We also release the data for future research in this direction