Graduation - 2025 - March (Embargoes)

Permanent URI for this collection

https://scholar.sun.ac.za/handle/10019.1/131921

Browse

Recent Submissions

Now showing 1 - 2 of 2
  • Thumbnail Image
    Item
    A computer vision framework towards automated scene understanding & analysis
    (2025-03) Sarah-lee de Greeff
    It is well-known that recent advancements in the domain of artificial intelligence and the increased capability of computer hardware have significantly advanced the field of computer vision – a field of study which enables computers to “see” and extract meaningful information from visual inputs, similar to human perception. A prominent application area within the domain of computer vision is scene understanding. Various powerful approaches towards scene understanding employ computer vision tasks to extrapolate semantic information about scenes, allowing computers to understand relationships between objects and their environments. Such computer vision tasks include object detection, recognition, tracking, pose estimation, and contextual reasoning. Most computer vision algorithms are deep learning based approaches but differ significantly in architecture. The computer vision tasks investigated in this thesis utilise architectures consisting of backbone, neck, and head architecture as well as alternative transformer architectures. Although computer vision applications are diverse, there remain fields that have not yet fully benefited from these developments. One such field is energy auditing – a process undertaken to evaluate and improve the energy management of buildings. In this thesis, a proof-of-concept framework is developed, capable of extracting information regarding appliances present in a given building scene or environment by employing object detection and object tracking tasks. The objective of the proposed framework is to train various object detection models and recommend the best-performing model for further implementation, in conjunction with object tracking models, to analyse video footage of environments needing to be audited. The framework facilitates the processing of raw data, training of object detection models with respect to the proposed data, and the deployment of the trained model with respect to unseen video footage. A structured literature review is conducted in this thesis to investigate the pertinent literature related to computer vision applications within the energy auditing domain. The fundamentals of deep learning, computer vision and energy auditing are also explored. The proposed framework is first applied to a subset of a publicly accepted benchmark dataset to verify its correct functioning. Subsequently, to further assess the framework’s performance and applicability, it is applied to a novel case study dataset provided by an industry partner, containing images of appliances common in an educational institution. The framework facilitates hyperparameter tuning to determine the best parameters for each model being trained. The best-performing model, RTDeTR, is then utilised to detect and track appliances of interest, providing information regarding the number of appliances present. The information attained by the models is essential for the environment’s energy consumption computation.
  • No Thumbnail Available
    Item
    Investigating the effects of chronic stress on cardiovascular and metabolic function
    (Stellenbosch University, 2025-03) Cairns, Megan Louise; Essop, Faadiel; Joseph, Danzil
    Cardiovascular diseases (CVD) represent a significant global health challenge, with ischemic heart disease being the foremost cause of morbidity and mortality. Recent studies identified mental illnesses and psychological stress as crucial risk factors contributing to the onset and progression of CVD. Despite a growing body of preclinical evidence linking stress to cardiovascular dysfunction, the specific pathophysiological mechanisms, particularly those that differ between sexes, remain inadequately understood. To investigate this knowledge gap, we conducted a study utilizing 10-week-old male and female Wistar rats, which were subjected to four weeks of daily 1-hour restraint stress versus controls. Following this intervention, behavioral tests and blood samples were collected to confirm successful induction of a chronic stress phenotype within the experimental cohorts. Rats were euthanized via sodium pentobarbitone overdose, and their isolated hearts subjected to 25 minutes of regional ischemia in an ex vivo setting. Cardiac function was assessed before and after reperfusion, together with molecular markers in non-ischemic and ischemic cardiac tissues post-reperfusion, while proteomic analyses focused on total and phosphorylated proteins in female post-reperfusion tissue. Our findings revealed distinct sex-specific responses to chronic restraint stress (CRS). Male rats exhibited significant reductions in circulating brain-derived neurotrophic factor (BDNF) and testosterone levels (both p<0.05), along with low-anxiety behavior and no changes in hypothalamic-pituitary axis hormones. In contrast, female rats subjected to CRS showed decreased plasma corticosterone levels (p<0.001), a reduced estradiol-to-progesterone ratio (p<0.01), and altered interleukin levels (IL-1α/1β/10/6; p<0.05). However, they also exhibited elevated cardiac troponin T (p<0.05), tumor necrosis factor-α (TNF-α; p<0.01), and adrenocorticotropic hormone (ACTH; p<0.01) compared to controls. These These changes collectively suggest a state of global cellular senescence and underlying cardiac damage. Despite these alterations in circulating markers, baseline and post-reperfusion cardiac functional parameters remained unchanged in isolated hearts across both sexes. However, post-ischemic recovery was significantly compromised in female rats, evidenced by attenuated work performance (p<0.05), stroke volume (p<0.05), and cardiac output (p<0.01) relative to controls and CRS males (p<0.05). Mitochondrial and redox assessments revealed an increase in fatty acid oxidation (p<0.05) and lipid peroxidation (p<0.05) within non-ischemic tissues of male rats, with decreased superoxide dismutase activity (p<0.05). Conversely, female mitochondria demonstrated resilience against chronic stress and ischemic injury despite notable lipid peroxidation (p<0.05) in non-ischemic tissues. Proteomic analysis unveiled significant alterations in proteins associated with cardiac contraction (e.g. desmin, titin, myosin light chains), mitochondrial biogenesis (e.g. TAMM41, MDH2, DHTDK1), protein homeostasis (e.g. DRAP1, EIF4ENIF1), cell death signaling pathways (e.g. ANAPC1, PPP1R13L, WSB2), and pro-fibrotic mediators (e.g. ACLY, COMP, TIMP3). These findings suggest that chronic stress induces substantial cardiac dysfunction particularly in females. Our research underscores that chronic stress can lead to sex-specific cardiac dysfunctions i.e. male rats adapting through metabolic and redox modifications, while females may exhibit heightened vulnerability due to a blunted hypothalamic-pituitary-adrenal axis response together with inflammatory changes. This study emphasizes the necessity for integrating sex-specific considerations into the prevention and treatment strategies for cardiovascular diseases to enhance therapeutic outcomes.