2024 3MT Contestants
The 3MT Competition highlights the breadth of graduate research undertaken at the University of Winnipeg. 2024's 3MT presentations were on topics diverse as conflict assessment, machine learning, Ghana’s chocolate industry, Bangladesh’s artisan weavers, seals, bats, resource management, soil quality, clean energy, media and politics, crime prevention, and empowering research practices.
Read about the contestants and their work below!
Janelle is an MSc candidate in the Bioscience, Technology, and Public Policy program.
Claws, Climate Change, and the Canadian Arctic: Studying Ringed Seal and Bearded Seal Diet
Ringed seals and bearded seals are at the top of the Arctic marine food chain and rely heavily on sea-ice for survival, making them vulnerable to rapid climate warming that is drastically changing Arctic ecosystems. Although both species forage in coastal habitats, bearded seals feed along the sea floor while ringed seals feed throughout the water column; setting the stage to investigate how each species will be impacted by continued environmental change. This project will examine dietary chemical tracers (stable isotopes and mercury) in seasonal layers present in seal claws to investigate how diets have changed over space and time in ringed and bearded seals in relation to changes in sea ice. This research will improve our ability to compare climate change impacts across species with different habitat use and life history strategies.
Olivia is a Master of Arts candidate in the Environmental and Social Change program.
Researcher Reflexivity on Empowering Practice in Community-Engaged Research
Empowerment is an essential practice of participatory research, however there are multiple and at times contradictory understandings of what ‘empowerment’ means. My thesis project explores definitions and applications of empowerment through interviews with experienced community-based researchers.
Pranav is an MSc candidate in the Bioscience, Technology, and Public Policy program.
Hibernating to Hibernate: How Do Bats Accumulate Fat to Prepare for Hibernation?
Hibernation significantly saves energy, but hibernators still need energy reserves to survive winter. Little brown bats are widespread hibernating bats in North America. They rely on fat accumulated in fall for hibernation, which is challenging to build up as the weather gets colder and insect prey scarcer. Despite these challenges, bats add 30-40% of their summer body mass as fat during only a few weeks but how they do this is unknown. I hypothesized that bats rely on short-term hibernation or torpor (decreased metabolism and body temperature) during fall and select cold roosts that maximize torpor. I used miniature radio-tags to record body temperature and roost preferences during fall 2022 and 2023. I found that bats “hibernate to hibernate”, regularly using torpor and roosting in cold roosts presumably to reserve energy for fat accumulation. This study improves understanding of hibernation biology and habitat requirements for this endangered species.
Shamim is an MA candidate in the Environmental and Social Change program.
Unlocking Green Energy Potential: A Policy Analysis of Biomass Energy in Manitoba
Bioenergy, derived from living organisms, specifically biomass, plays a vital role in energy and fuel production. Despite its potential, Canada only generates 6% of its energy from bioenergy, highlighting the need for effective policies to boost its share in the energy mix. This research focuses on Manitoba, where the Clean Energy Strategy emphasizes the substantial cost-saving potential of the biomass heat industry. The study aims to investigate Manitoba's sustainable forest management policies, energy-related legislation, and strategies to identify policy requirements. Using the Policy Mix Analysis Framework, the research will analyse existing policies to understand their support for biomass energy development. Employing a case study methodology, the research will triangulate findings through document analysis, interviews, and field observation. The outcomes of this study will provide insights for policymakers, researchers, and industry representatives, facilitating a better understanding of the policy landscape's impact on biomass energy development. Supporting biomass-based bioenergy is crucial for achieving Canada's net-zero goal by 2050, aligning with environmental, economic, and social objectives.
Ershiya is an MSc candidate in the Bioscience, Technology, and Public Policy program.
Summer Roost Selection of Endangered Little Brown Bats
Conservation of endangered wild animals relies on protecting the habitats essential for their survival and reproduction, but for many species, where these habitats are or how animals select them remain unknown. In Canada, little brown bats (Myotis lucifugus) are endangered, yet our understanding of the roosts they use to raise their pups is limited, posing challenges for conservation efforts. The main question of my thesis is: Why do bats choose certain roosts? I predicted that they would select unshaded, south-facing structures near freshwater sources to maintain warmth and reduce travel costs to feeding and drinking areas. During the summers of 2021 and 2022, I radio-tagged 30 bats from three maternity colonies near Kenora, Ontario, and tracked them for about a month each year. As predicted, the bats selected south-facing roosts closer to freshwater than the average structure in the study area. This highlights the importance of prioritizing proximity to freshwater sources to help bat populations recover.
Jaher is an MSc candidate in the Applied Computer Science and Society program.
Exploring Histogram Equalization Technique in Facial Emotion Recognition Datasets with Convolutional Neural Networks
Emotions play a significant role in human interactions, serving as essential mediators in social communication systems. Human expression of emotions incorporates diverse modalities, such as facial expressions and speech patterns. Automatic facial emotion recognition (AFER) is a computer vision task that has gained popularity in areas such as multimodal sentimental analysis and depression detection. AFER aims to detect and classify human emotions using machine learning and deep neural network (DNN) algorithms. This research leverages pre-trained DNN models (VGG 16 and 19) in detecting seven emotions: anger, fear, sadness, surprise, neutral/contempt, and disgust from two benchmark datasets. Our work focuses on two main tasks: enhancing the quality of the images via histogram equalization and classifying emotions via pre-trained DNN models. Model tuning and parameter optimization were performed on the two models. The obtained results were very promising in terms of best classification accuracies: 94.21% (KDEF) & 98% ( CK+ ) datasets.
Aishika is an MSc candidate in the Environmental and Social Change program.
Phosphorus Speciation Changes with the Application of Single and Blended Soil Amendments
Phosphorus (P) is an essential nutrient for plant growth and one of the primary limiting nutrients in freshwater ecosystems for algal growth. Phosphorus released from agricultural soils is a major source of P to surface water bodies, resulting in eutrophication. Soil amendments are effective in mitigating P loss from soils, however, the mechanisms by which P loss is reduced are poorly understood. This study aims to evaluate the solubility and transformations of soil P with the application of single or blended amendments at varying rates under different moisture status and temperatures. To accomplish my objectives of the research, I will conduct multiple laboratory incubation experiments using natural soils and model soils (artificial soils). The results of this study will help to identify amendment blends that could potentially be used to mitigate P losses from soils under natural environments by retaining P in soils.
Indeera is an MSc candidate in the Environmental and Social Change program.
Blended Amendments: A Recipe to Save Our Lakes
Phosphorus (P) is a crucial element in agroecosystems; however, its export through watershed runoff can rapidly increase the algal growth in receiving water bodies and result in water quality deterioration and fish kills. Snowmelt runoff is the main pathway by which agricultural P is transported to water bodies in the Canadian prairies. Prolonged snowmelt flooding on the soil surface facilitates the P mobilization into the snowmelt water. Chemical soil amendments can reduce the migration of P from the soil to the aquatic environment, but their effects under cold environments are less studied. This study investigates the effect of blended amendments of gypsum (CaSO4·2H2O), alum [Al2(SO4)3·18H2O], ferric chloride (FeCl3), and magnesium sulfate (MgSO4) on phosphorus loss to snowmelt from manured soils. The results of this research would help to mitigate P loss to waterways using soil amendment blend in high legacy P soils in Manitoba.
Ashique is an MSc candidate in the Applied Computer Science and Society program.
Dataset Optimization Using Image Processing
Effective Machine Learning and Convolutional Neural Network model training often demand extensive datasets. However, redundancy within datasets, where similar or identical images offer limited useful or informative content, becomes a prevalent challenge. This redundancy not only escalates resource and computational power consumption but also hampers dataset sharing among researchers. In our study, we implemented the average hashing algorithm to identify similarity among plant images, leading to the removal of noisier and low-resolution redundant images and the creation of new refined datasets based on the score of that algorithm. Subsequently, we used all these original and distilled datasets to train different instances of the same supervised learning model. Our proposed approach demonstrated the model's accuracy stability up to a certain similarity score, with minor fluctuations on two different testing datasets. Beyond this threshold, a substantial decline in accuracy was observed.
Nowsheen is an MDP candidate in the Master's in Development Practice: Indigenous Development program.
Spinning History: How Colonization Unraveled Bangladesh's Muslin Mastery
Weavers in Dhaka, Bangladesh, used to make this incredibly fine cloth using a method called the discontinuous weft technique. This technique required the weaver to work two weft layers – one as fine as spider silk to hold the cloth together and the other forming the pattern. “Muslin,” or as poets of ancient Mughal Darbar used to call it- “Baft Hawa,” is the lost pride, which was once a chief attraction of Bengal. This name translates to “woven air” to depict the lightness and premium feel of the fabric, which was most often seen in clothes worn by the nobles.
The reason behind this lost original muslin is the tedious and long yet skilled process of weaving Muslin, lack of branding over contemporary media, extreme political turmoil, colonization, and economic unrest in Bangladesh. This thesis aims to rebrand Muslin and Bangladeshi Artisans globally.
Muhammad is an MSc candidate in the Applied Computer Science and Society Program.
Unveiling the Unseen: Advancing Memory Forensics in Redis for Robust Database Security
This study explores the intersection of databases and memory forensics, focusing on challenges presented by Redis, an in-memory database system. It aims to address critical issues in digital forensics, such as the detection of unlogged actions in Database Management Systems (DBMS), the adaptability of memory forensics to evolving data structures, and the reliability of linking memory events to logs. Using Redis as a case study, this research employs innovative methodologies to systematically analyze Redis commands for vulnerability detection and to bridge the gap between memory forensics and Redis log analysis. Preliminary findings indicate improved precision and comprehensiveness in database forensic investigations, offering practical insights and tools for addressing real-world DBMS security concerns. The study contributes to the advancement of forensic approaches in the context of in-memory databases, highlighting the importance of integrating memory forensics with log analysis for enhanced security and forensic capabilities.
Shaghayegh is an MA candidate in the Applied Economics program.
Social Return on Investment
In Canada, pressing social issues like homelessness, unemployment, and rising crime rates demand attention. To address these, social recreational programs offer a promising solution, especially when analyzed through Social Return on Investment (SROI). Unlike traditional cost-benefit analysis, SROI evaluates intangible benefits, like health and kindness, revealing significant societal gains. For instance, a study on a boxing program showed a staggering return of at least 91 dollars for every dollar invested, highlighting the immense potential of such initiatives. As an economist, leveraging data-driven approaches like SROI can guide decision-making, ensuring optimal allocation of resources. This underscores the importance of investing in simple yet impactful social activities, as they can yield substantial returns and address complex societal challenges effectively.
Mawulom is an MA candidate in the Applied Economics program.
Chocolate: The Perfect Blend of Cocoa and Economic Institutions
In 2022 the chocolate industry was worth $200 billion. The main ingredient in chocolate, cocoa, is one of Ghana’s top exports. In fact, 60% of the world’s cocoa comes from farms in Ghana and Côte d'Ivoire. Ironically, neither country has a reputation or industry in chocolate production.
Chocolate is lucrative, and Ghana has cocoa in abundance. So why is Ghana’s chocolate industry not dominant? My study seeks to answer this research question. I studied the historical economics of the cocoa trade, domestic and international policies of the sector, and the cost-benefit of processing cocoa in Ghana relative to leading chocolate producers in Switzerland and the US.
My research reveals that while natural resources are important, they must be augmented by enabling systems that incentivize innovation. This study provides an innovative case study on how governments can support markets in contexts where natural resources are abundant – a case study of industrial policy.
Fran is an MSc candidate in the Bioscience, Technology, and Public Policy Program.
Turn Down for WAT: Breast Milk Feeding Promotes Fat Breakdown
Components of breast milk work together to promote healthy development. Milk nanovesicles are molecules in breast milk that influence growth by transporting genetic information from moms to babies. However, maternal diet influences breastmilk composition. Babies who are born to mothers with obesity have dysregulated fat cells, which elevates the risk of developing metabolic issues. Obesity is a state of excess fat (lipid) storage and decreased fat breakdown (lipolysis), which leads to cellular stress and pro-inflammation. The peroxisome proliferator-activated receptors (PPARs) address this imbalance. PPARs activate genes related to the uptake, transport, and storage of lipids. Babies who are born to mothers with obesity have impaired PPARs. Yet, breast milk feeding increases lipolysis and improves metabolic health outcomes. My study aims to understand how breast milk feeding confers these benefits, and specifically, how milk nanovesicles influence PPARs to promote lipolysis. My study will further our understanding of the protective effects of breastmilk and provide insight into strategies to mitigate developing metabolic disease.
Orthy is an MA candidate in the Peace and Conflict Studies program.
Social Prism of Conflict: A Conflict Analysis Model
Conflict is a natural and normal part of human existence. To transform these conflicts into sustainable peace initiatives identifying the root causes is crucial. Therefore, this analysis model considered the multiple conflict spectrums analysis to develop organic peacebuilding approaches.
Yujie is an MSc candidate in the Bioscience, Technology, and Public Policy program.
Optimal Design of Wireless Radiofrequency (RF) Head Coil at 1.5T IntraOperative MRI System
The RF system is one of the core components that excites/receives the magnetic resonance signal, and its performance directly influences MRI quality. The previous RF coil designed by our team was used for animal tests which showed significant benefits with high signal-to-noise (SNR) as well as simple structure and low cost. In the project that I'm doing, a new head RF coil with high SNR will be designed using optimal design method for application in clinical imaging procedures. Electromagnetic (EM) simulation will be used for evaluation of EM-field uniformity. Phantom (ex vivo) and human brain (in vivo) experiments will be applied to validate MR imaging quality and clinical application. This project will provide innovative technology for the diagnosis of brain diseases and the study of brain lesions.
Marcos is an MA candidate in the Criminal Justice program.
Panic Devils and Media Morals
Building on Stanley Cohen’s work, I explore this assertion by sociolegal experts Malcolm Feeley and Jonathan Simon:
There is now a vast body of research that supports this reformulation of Cohen’s moral panic thesis…newer literature treats them as ‘actors’ who can not only amplify crises but also create the events that precipitate this sense of heightened urgency. Thus, rather than just seizing the moment when disasters occur, they have the capacity to manufacture disasters from whole cloth and the sense of accompanying moral outrage…policy entrepreneurs have learned not only how to mobilize and amplify moral outrage in order to generate moral panics, but also how to create the conditions that give rise to it in the first place. Moral panics have become institutionalized, transformed from ad hoc, episodic, and short-lived occurrences into regularly orchestrated campaigns; indeed, into a permanent condition of personal insecurity. (Feeley & Simon p. 43, 2007).
Jessie is an MA candidate in the Cultural Studies program.
Dancing with Danger: Is the Nightclub a Site of Struggle for Freedom from Oppression?
The contemporary nightclub is often associated with popular youth culture, stirring up images of fun and freedom; the freedom to explore sexual and gender expression, freedom of a person’s body through dance, and the letting go of inhibition with the use of substances. At the same time, however, the nightclub is also widely known for being a place where people (particularly women and queer people) are exposed to both threatened and actual violence that can result not only in lasting psychological damage but the loss of life. Cultural studies theorist Stuart Hall asserts that popular culture is best understood as a site of struggle for power and liberation. From a cultural studies lens, how might we come to understand the nightclub an important cultural site of struggle for women and queer people against systems of oppressive power?