Showing items 1 - 20 of 35

Computer Science Education in Selected Countries from Sub-Saharan Africa

- Bainomugisha, Engineer, Bradshaw, Karen L, Ujakpa, Martin Mabeifam, Nakatumba-Nabende, Joyce, Nderu, Lawrence, Mduma, Neema, Kihoza, Patrick, Irungu, Annette

  • Authors: Bainomugisha, Engineer , Bradshaw, Karen L , Ujakpa, Martin Mabeifam , Nakatumba-Nabende, Joyce , Nderu, Lawrence , Mduma, Neema , Kihoza, Patrick , Irungu, Annette
  • Date: 2024
  • Subjects: To be catalogued
  • Language: English
  • Type: text , article
  • Identifier: http://hdl.handle.net/10962/440223 , vital:73758 , xlink:href="https://doi.org/10.1145/3643037"
  • Description: Computer Science education in sub-Saharan Africa has evolved over the past decades. The number of institutions offering distinct undergraduate programs has grown, thus increasing the number of students enrolling in the computer science discipline. Several computer science degree programs have emerged with one of the objectives being to satisfy the growing demand for local talent and skills. In this paper, we provide a snapshot of the evolution of undergraduate computer science education in selected countries in Sub-Saharan Africa over the past 20+ years and an overview of the developments in computer science education and observed trends. The setup of educational institutions in Africa and the operational context requires unique modalities for the design and delivery of computer science education that meets the demands of the industry, amongst others. This paper provides insights into the best practices in the computer science curricula in the selected countries, as well as an overview of the pedagogical and delivery approaches to computer science education. The paper highlights case studies from institutions in the selected countries, namely Uganda, South Africa, Ghana, Tanzania, and Kenya with a consolidated summary of the current and emerging challenges and opportunities in all these countries. The paper concludes by providing persectives on the future landscape of computer science in Sub-Saharan Africa.Computer Science Education in Selected Countries from Sub-Saharan AfricaBy Engineer Bainomugisha, Makerere University, Uganda, Karen Bradshaw, Rhodes University, South Africa, Martin Mabeifam Ujakpa, Ghana Communication Technology University, Ghana, Joyce Nakatumba-Nabende, Makerere University, Uganda, Lawrence Nderu, Jomo Kenyatta University of Agriculture and Technology, Kenya, Neema Mduma, Nelson Mandela African Institution of Science and Technology, Tanzania, Patrick Kihoza, Mzumbe University, Tanzania and Annette Irungu, Jomo Kenyatta University of Agriculture and Technology, Kenya
  • Full Text:
  • Date Issued: 2024

Data Visualization of Budgeting Assumptions: An Illustrative Case of Trans-disciplinary Applied Knowledge

- Cuthbert, Carol E, Pearse, Noel J, Bradshaw, Karen L

  • Authors: Cuthbert, Carol E , Pearse, Noel J , Bradshaw, Karen L
  • Date: 2024
  • Subjects: To be catalogued
  • Language: English
  • Type: text , article
  • Identifier: http://hdl.handle.net/10962/440236 , vital:73759 , xlink:href="https://doi.org/10.54808/JSCI.22.01.130"
  • Description: Trans-disciplinary research combines different fields into new conceptual and methodological frameworks. In this study, the SECI model of knowledge creation, which consists of Socialization, Externalization, Combination, and Internalization conversion modes, is used to analyze the implementation of a structured budgeting visualization system by a trans-disciplinary team. Through applied research in implementing a global budgeting system, budgeting assumptions are made explicit through visualization, transforming the approach to the budgeting process and its accuracy. This visualization, in turn, is enabled by assumptions underlying revenue planning, business services and employee compensation, and a visual process. The system displays a stepped approach, indicated by icons, representing the tasks involved in the budget process. For example, the system requires uploading the previous year’s information, setting the assumptions, calculating the suggested figures based on assumptions, and amending the proposed outcome. As adapted by Rice and Rice (2005), SECI is applied as the socialization of tacit-to-tacit budgeting assumption knowledge is solidified during the design phase of this transformation exercise. The externalization phase, in which budgeting assumptions are transformed from tacit to explicit, is evidenced during the configuration phase of the new system. The systemic collaboration results in the explicit assumptions being collectively leveraged across the regions during and after the “go-live” phase of system development. Finally, the internalization phase involves the explicit assumptions being transformed into new tacit knowledge as the experts evolve new assumptions derived from the transformation process. Semiotics provides variance information through hue, with, for example, darker colours indicating higher variances. This trans-disciplinary communication provides the means for increased efficiency and effectiveness. The resulting budget framework is visually validated through a heatmap by comparing the budgeting accuracy and assumption complexity between the different regions where it was implemented. In summary, value is added in developing a new data visualization process, focusing on the role of budgeting assumptions and using planning process visualizations. This approach improves communication efficiency, effectiveness, and understanding of budgeting while enhancing accuracy.
  • Full Text:
  • Date Issued: 2024

Agent-based model development of a complex socio-ecological system: Methods for overcoming data and domain limitations

- James, C. L, Bradshaw, Karen L

  • Authors: James, C. L , Bradshaw, Karen L
  • Date: 2023
  • Subjects: To be catalogued
  • Language: English
  • Type: text , article
  • Identifier: http://hdl.handle.net/10962/440189 , vital:73755 , xlink:href="https://doi.org/10.1016/j.ecoinf.2023.102224"
  • Description: Agent-based models (ABMs) are appropriate tools for modelling socio-ecological systems due to their ability to handle complexity. However, development of such models is often an intensive process. There are many tutorials on the general methods and steps in ABM development but there are not necessarily practical details on how to overcome certain challenges. Honeybush (Cyclopia spp.), a kind of fynbos vegetation found in the Western and Eastern Cape mountains, is an important ecological and agricultural product in South Africa. It is considered a complex system due to its variability and unpredictability. The honeybush tea industry faces the challenge of meeting emerging market demands while maintaining sustainable harvesting practices, in the midst of an uncertain future due to climate change. We created a prototype model, HoneybushModel, using the MARS framework, a C# multi-agent simulation toolkit. The model was validated using historic data. Whilst outlining the development processes used to create the Honeybush Model, this paper provides a methodology for the development of such models and demonstrates techniques for addressing data, domain and framework limitations. The implemented model also acts as a case study for similar systems that could be modelled using an ABM.
  • Full Text:
  • Date Issued: 2023

Agent-Based Modeling and Simulation for Transmission Dynamics and Surveillance of Dengue: Conceptual and Design Model

- Pascoe, Luba, Nyambo, Devotha G, Bradshaw, Karen L, Clemen, Thomas

  • Authors: Pascoe, Luba , Nyambo, Devotha G , Bradshaw, Karen L , Clemen, Thomas
  • Date: 2023
  • Subjects: To be catalogued
  • Language: English
  • Type: text , article
  • Identifier: http://hdl.handle.net/10962/440200 , vital:73756 , xlink:href="https://doi.org/10.1109/AFRICON55910.2023.10293299"
  • Description: African countries need to strengthen surveillance and control of arboviral diseases such as dengue due to increased outbreaks and spread of arboviruses. Climatic, socio-environment, and ecological variables influence the spread of dengue fever in Sub-Saharan Africa. This paper presents an Agent-Based conceptual and design model for dengue fever developed using the Multi-Agent Research and Simulation (MARS) framework. The study analyzes dengue fever's spatial distribution and identifies the causal relationship between the disease and its climatic and environmental variables. Agent-based modeling (ABM) was used to comprehend the spatial patterns of variation to determine the ecological association between the observed spatio-temporal variations in dengue fever. The domain and design model of an ABM for the surveillance of dengue fever is presented based on the Overview, Design Concepts, and Details (ODD) protocol. Model input parameters and input data for the study area are also presented. The dengue ABM can be adopted and reused for modeling other diseases and other complex problems from different domains while ensuring that their unique characteristics and appropriate modifications are considered to ensure the model's validity and relevance to the new context.
  • Full Text:
  • Date Issued: 2023

An Exploration of Flow Control Using Machine Learning and Computational Fluid Dynamics

- Bradshaw, Karen L, Cornfield Matthew

  • Authors: Bradshaw, Karen L , Cornfield Matthew
  • Date: 2023
  • Subjects: To be catalogued
  • Language: English
  • Type: text , article
  • Identifier: http://hdl.handle.net/10962/440211 , vital:73757 , xlink:href="https://doi.org/10.59200/ICARTI.2023.017"
  • Description: Although numerous studies relating to computational fluid dynamics and machine learning have been conducted in relation to automotive development, the majority focus on either early development using completed 3D models, or the final testing stages of development, or machine learning accelerated computational fluid dynamic simulations. While this approach is helpful in software development and simulation, it is not easily adaptable to automotive design where the final model is constantly changing and being modified. Consequently, the aim of this study is to propose a method for conducting computational fluid dynamics and machine learning concurrently to accelerate the development process. The proposed method is used to design and improve the aerodynamic efficiency of an object. The approach focuses on developing, implementing, and comparing machine learning models capable of generating optimised three-dimensional objects with the required geometry to direct airflow paths required in applications such as pressure generation, as needed for both active and passive flow control. The study concludes that both decision tree regression and long short-term memory (LSTM) autoencoder models could be used to optimise the aerodynamic efficiency of solid bodies, but that the LSTM autoencoder performs better overall. An undesirable effect of the shape optimisation is an overall reduction in shape size as optimization increases.
  • Full Text:
  • Date Issued: 2023

Deep Palmprint Recognition with Alignment and Augmentation of Limited Training Samples

- Brown, Dane L, Bradshaw, Karen L

  • Authors: Brown, Dane L , Bradshaw, Karen L
  • Date: 2022
  • Subjects: To be catalogued
  • Language: English
  • Type: text , article
  • Identifier: http://hdl.handle.net/10962/440249 , vital:73760 , xlink:href="https://doi.org/10.1007/s42979-021-00859-3"
  • Description: This paper builds upon a previously proposed automatic palmprint alignment and classification system. The proposed system was geared towards palmprints acquired from either contact or contactless sensors. It was robust to finger location and fist shape changes—accurately extracting the palmprints in images without fingers. An extension to this previous work includes comparisons of traditional and deep learning models, both with hyperparameter tuning. The proposed methods are compared with related verification systems and a detailed evaluation of open-set identification. The best results were yielded by a proposed Convolutional Neural Network, based on VGG-16, and outperforming tuned VGG-16 and Xception architectures. All deep learning algorithms are provided with augmented data, included in the tuning process, enabling significant accuracy gains. Highlights include near-zero and zero EER on IITD-Palmprint verification using one training sample and leave-one-out strategy, respectively. Therefore, the proposed palmprint system is practical as it is effective on data containing many and few training examples.
  • Full Text:
  • Date Issued: 2022

Deep palmprint recognition with alignment and augmentation of limited training samples

- Brown, Dane L, Bradshaw, Karen L

  • Authors: Brown, Dane L , Bradshaw, Karen L
  • Date: 2022
  • Subjects: To be catalogued
  • Language: English
  • Type: text , article
  • Identifier: http://hdl.handle.net/10962/464074 , vital:76473 , xlink:href="https://doi.org/10.1007/s42979-021-00859-3"
  • Description: This paper builds upon a previously proposed automatic palmprint alignment and classification system. The proposed system was geared towards palmprints acquired from either contact or contactless sensors. It was robust to finger location and fist shape changes—accurately extracting the palmprints in images without fingers. An extension to this previous work includes comparisons of traditional and deep learning models, both with hyperparameter tuning. The proposed methods are compared with related verification systems and a detailed evaluation of open-set identification. The best results were yielded by a proposed Convolutional Neural Network, based on VGG-16, and outperforming tuned VGG-16 and Xception architectures. All deep learning algorithms are provided with augmented data, included in the tuning process, enabling significant accuracy gains. Highlights include near-zero and zero EER on IITD-Palmprint verification using one training sample and leave-one-out strategy, respectively. Therefore, the proposed palmprint system is practical as it is effective on data containing many and few training examples.
  • Full Text:
  • Date Issued: 2022

Drowning in data, thirsty for information and starved for understanding: A biodiversity information hub for cooperative environmental monitoring in South Africa

- MacFadyen, Sandra, Allsopp, Nicky, Altwegg, Res, Archibald, Sally, Botha, Judith, Bradshaw, Karen L, Carruthers, Jane, De Klerk, Helen, de Vos, Alta, Distiller, Greg, Foord, Stefan, Freitag-Ronaldson, Stefanie, Gibbs, Richard, Hamer, Michelle, Landi, Pietro, MacFayden, Duncan, Manuel, Jeffrey, Midgley, Guy, Moncrieff, Glenn, Munch, Zahn, Mutanga, Onisimo, Sershen, Nenguda, Rendani, Ngwenya, Mzabalazo, Parker, Daniel M, Peel, Mike, Power, John, Pretorius, Joachim, Ramdhani, Syd, Robertson, Mark P, Rushworth, Ian, Skowno, Andrew, Slingsby, Jasper, Turner, Andrew, Visser, Vernon, van Wageningen, Gerhard, Hui, Cang

Review of Importance of Weather and Environmental Variables in Agent-Based Arbovirus Models

- Pascoe, Luba, Clemen, Thomas, Bradshaw, Karen L, Nyambo, Devotha G

  • Authors: Pascoe, Luba , Clemen, Thomas , Bradshaw, Karen L , Nyambo, Devotha G
  • Date: 2022
  • Subjects: To be catalogued
  • Language: English
  • Type: text , article
  • Identifier: http://hdl.handle.net/10962/440300 , vital:73764 , xlink:href="https://doi.org/10.3390/ijerph192315578"
  • Description: The study sought to review the works of literature on agent-based modeling and the influence of climatic and environmental factors on disease outbreak, transmission, and surveillance. Thus, drawing the influence of environmental variables such as vegetation index, households, mosquito habitats, breeding sites, and climatic variables including precipitation or rainfall, temperature, wind speed, and relative humidity on dengue disease modeling using the agent-based model in an African context and globally was the aim of the study. A search strategy was developed and used to search for relevant articles from four databases, namely, PubMed, Scopus, Research4Life, and Google Scholar. Inclusion criteria were developed, and 20 articles met the criteria and have been included in the review. From the reviewed works of literature, the study observed that climatic and environmental factors may influence the arbovirus disease outbreak, transmission, and surveillance. Thus, there is a call for further research on the area. To benefit from arbovirus modeling, it is crucial to consider the influence of climatic and environmental factors, especially in Africa, where there are limited studies exploring this phenomenon.
  • Full Text:
  • Date Issued: 2022

Shrub Detection in High-Resolution Imagery: A Comparative Study of Two Deep Learning Approaches

- James, Katherine M F, Bradshaw, Karen L

  • Authors: James, Katherine M F , Bradshaw, Karen L
  • Date: 2022
  • Subjects: To be catalogued
  • Language: English
  • Type: text , book
  • Identifier: http://hdl.handle.net/10962/440326 , vital:73766 , ISBN 9783030955021 , https://doi.org/10.1007/978-3-030-95502-1_41
  • Description: A common task in high-resolution remotely-sensed aerial imagery is the detection of particular target plant species for various ecological and agricultural applications. Although traditionally object-based image analysis approaches have been the most popular method for this task, deep learning approaches such as image patch-based convolutional neural networks (CNNs) have been seen to outperform these older approaches. To a lesser extent, fully convolutional networks (FCNs) that allow for semantic segmentation of images, have also begun to be used in the broader literature. This study investigates patch-based CNNs and FCN-based segmentation for shrub detection, targeting a particular invasive shrub genus. The results show that while a patch-based CNN demonstrates strong performance on ideal image patches, the FCN outperforms this approach on real-world proposed image patches with a 52% higher object-level precision and comparable recall. This indicates that FCN-based segmentation approaches are a promising alternative to patch-based approaches, with the added advantage of not requiring any hand-tuning of a patch proposal algorithm.
  • Full Text:
  • Date Issued: 2022

Mapping Computational Thinking Skills to the South African Secondary School Mathematics Curriculum

- Bradshaw, Karen L, Milne, Shannon

  • Authors: Bradshaw, Karen L , Milne, Shannon
  • Date: 2021
  • Subjects: To be catalogued
  • Language: English
  • Type: text , book
  • Identifier: http://hdl.handle.net/10962/440285 , vital:73763 , ISBN 9783030950033 , https://doi.org/10.1007/978-3-030-95502-1_41
  • Description: Computational thinking (CT) is gaining recognition as an important skill for learners in both Computer Science (CS) and several other disciplines, including mathematics. In addition, researchers have shown that there is a direct correlation between poor mathematical skills and the high attrition rate of CS undergraduates. This research investigates the use of nine core CT skills in the South African Grades 10–12 Mathematics curriculum by mapping these skills to the objectives given in each of the topics in the curriculum. The artefact developed shows that all the identified CT skills are used in the curriculum. With the use of this mapping, future research on interventions to develop these skills through mathematics at secondary school, should produce school leavers with better mathematical and problem solving abilities, which in turn, might contribute to better success rates in CS university courses.
  • Full Text:
  • Date Issued: 2021

Segmentation of Tuta Absoluta’s Damage on Tomato Plants: A Computer Vision Approach

- Loyani, Loyani K, Bradshaw, Karen L, Machuze, Dina

  • Authors: Loyani, Loyani K , Bradshaw, Karen L , Machuze, Dina
  • Date: 2021
  • Subjects: To be catalogued
  • Language: English
  • Type: text , article
  • Identifier: http://hdl.handle.net/10962/440313 , vital:73765 , xlink:href="https://doi.org/10.1080/08839514.2021.1972254"
  • Description: Tuta absoluta is a major threat to tomato production, causing losses ranging from 80% to 100% when not properly managed. Early detection of T. absoluta’s effects on tomato plants is important in controlling and preventing severe pest damage on tomatoes. In this study, we propose semantic and instance segmentation models based on U-Net and Mask RCNN, deep Convolutional Neural Networks (CNN) to segment the effects of T. absoluta on tomato leaf images at pixel level using field data. The results show that Mask RCNN achieved a mean Average Precision of 85.67%, while the U-Net model achieved an Intersection over Union of 78.60% and Dice coefficient of 82.86%. Both models can precisely generate segmentations indicating the exact spots/areas infested by T. absoluta in tomato leaves. The model will help farmers and extension officers make informed decisions to improve tomato productivity and rescue farmers from annual losses.
  • Full Text:
  • Date Issued: 2021

A Critical Evaluation of Validation Practices in the Forensic Acquisition of Digital Evidence in South Africa

- Jordaan, Jason, Bradshaw, Karen L

  • Authors: Jordaan, Jason , Bradshaw, Karen L
  • Date: 2020
  • Subjects: To be catalogued
  • Language: English
  • Type: text , book
  • Identifier: http://hdl.handle.net/10962/440174 , vital:73754 , ISBN 9783030660390 , https://doi.org/10.1007/978-3-030-66039-0_9
  • Description: Accepted digital forensics practice requires the tools used in the forensic acquisition of digital evidence to be validated, meaning that the tools perform as intended. In terms of Sect. 15 of the Electronic Communications and Transactions Act 25 of 2002 in South Africa, validation would contribute to the reliability of the digital evidence. A sample of digital forensic practitioners from South Africa was studied to determine to what extent they make use of validated forensic tools during the acquisition process, and how these tools are proven to be validated. The research identified significant concerns, with no validation done, or no proof of validation done, bringing into question the reliability of the digital evidence in court. It is concerning that the justice system itself is not picking this up, meaning that potentially unreliable digital evidence is used in court.
  • Full Text:
  • Date Issued: 2020

Detecting plant species in the field with deep learning and drone technology:

- James, Katherine M F, Bradshaw, Karen L

  • Authors: James, Katherine M F , Bradshaw, Karen L
  • Date: 2020
  • Language: English
  • Type: text , article
  • Identifier: http://hdl.handle.net/10962/160445 , vital:40446 , https://0-doi.org.wam.seals.ac.za/10.1111/2041-210X.13473
  • Description: Aerial drones are providing a new source of high‐resolution imagery for mapping of plant species of interest, amongst other applications. On‐board detection algorithms could open the door to allow for applications in which drones can intelligently interact with their environment. However, the majority of plant detection studies have focused on detection in post‐flight processed orthomosaics. Greater research into developing detection algorithms robust to real‐world variations in environmental conditions is necessary, such that they are suitable for deployment in the field under variable conditions. We outline the steps necessary to develop such a system, show by example how real‐world considerations can be addressed during model training and briefly illustrate the performance of our best performing model in the field when integrated with an aerial drone.
  • Full Text:
  • Date Issued: 2020

Detecting Similarity in Multi-procedure Student Programs Using only Static Code Structure

- Bradshaw, Karen L, Chindeka, Vongai

  • Authors: Bradshaw, Karen L , Chindeka, Vongai
  • Date: 2020
  • Subjects: To be catalogued
  • Language: English
  • Type: text , book
  • Identifier: http://hdl.handle.net/10962/440260 , vital:73761 , ISBN 9783030356286 , https://doi.org/10.1007/978-3-030-35629-3_14
  • Description: Plagiarism is prevalent in most undergraduate programming courses, including those where more advanced programming is taught. Typical strategies used to avoid detection include changing variable names and adding empty spaces or comments to the code. Although these changes affect the visual components of the source code, the underlying structure of the code remains the same. This similarity in structure can indicate the presence of plagiarism. A system has been developed to detect the similarity in the structure of student programs. The detection system works in two phases: The first phase parses the source code and creates a syntax tree, representing the syntactical structure of each of the programs, while the second takes as inputs two program syntax trees and applies various comparison algorithms to detect their similarity. The outcome of the comparison allows the system to report a result from one of four similarity categories: identical structure, isomorphic structure, containing many structural similarities, and containing few structural similarities. Empirical tests on small sample programs show that the prototype implementation is effective in detecting plagiarism in source code, although in some cases manual checking is needed to confirm the presence of plagiarism.
  • Full Text:
  • Date Issued: 2020

Environmental health promotion at a National Science Festival: An experiential-education based approach

- Duxbury, Theodore O, Bradshaw, Karen L, Khamanga, Sandile, Tandlich, Roman, Srinivas, Sunitha C

  • Authors: Duxbury, Theodore O , Bradshaw, Karen L , Khamanga, Sandile , Tandlich, Roman , Srinivas, Sunitha C
  • Date: 2020
  • Subjects: To be catalogued
  • Language: English
  • Type: text , article
  • Identifier: http://hdl.handle.net/10962/440274 , vital:73762 , xlink:href="https://doi.org/10.1080/1533015X.2019.1567406"
  • Description: To increase individual and communal environmental health awareness through an experiential-education project. A computer-based pre- and postintervention quiz; an educational poster; an interactive board game; and a take-home information leaflet were utilized for a school learners-centered health promotion exhibit at a National Science Festival in Grahamstown, South Africa. Out of all the participants, 55.7% were female, and 76.5% attended or had attended a government school. Participants showed significant improvement in their pre- and post educational intervention. The exhibit was effective in educating participants on environmental health, natural resources, and the impact environmental pollution has on their quality of life.
  • Full Text:
  • Date Issued: 2020

Design and evaluation of bulk data transfer extensions for the NFComms framework

- Bradshaw, Karen L, Irwin, Barry V W, Pennefather, Sean

  • Authors: Bradshaw, Karen L , Irwin, Barry V W , Pennefather, Sean
  • Date: 2019
  • Subjects: To be catalogued
  • Language: English
  • Type: text , article
  • Identifier: http://hdl.handle.net/10962/430369 , vital:72686 , https://hdl.handle.net/10520/EJC-1d75c01e79
  • Description: We present the design and implementation of an indirect messaging extension for the existing NFComms framework that provides communication between a network flow processor and host CPU. This extension addresses the bulk throughput limitations of the framework and is intended to work in conjunction with existing communication mediums. Testing of the framework extensions shows an increase in throughput performance of up to 268 that of the current direct message passing framework at the cost of increased single message latency of up to 2. This trade-off is considered acceptable as the proposed extensions are intended for bulk data transfer only while the existing message passing functionality of the framework is preserved and can be used in situations where low latency is required for small messages.
  • Full Text:
  • Date Issued: 2019

Improved palmprint segmentation for robust identification and verification

- Brown, Dane L, Bradshaw, Karen L

  • Authors: Brown, Dane L , Bradshaw, Karen L
  • Date: 2019
  • Subjects: To be catalogued
  • Language: English
  • Type: text , article
  • Identifier: http://hdl.handle.net/10962/460576 , vital:75966 , xlink:href="https://doi.org/10.1109/SITIS.2019.00013"
  • Description: This paper introduces an improved approach to palmprint segmentation. The approach enables both contact and contactless palmprints to be segmented regardless of constraining finger positions or whether fingers are even depicted within the image. It is compared with related systems, as well as more comprehensive identification tests, that show consistent results across other datasets. Experiments include contact and contactless palmprint images. The proposed system achieves highly accurate classification results, and highlights the importance of effective image segmentation. The proposed system is practical as it is effective with small or large amounts of training data.
  • Full Text:
  • Date Issued: 2019

Linking scales and disciplines: an interdisciplinary cross-scale approach to supporting climate-relevant ecosystem management

- Berger, Christian, Bieri, Mari, Bradshaw, Karen L, Brümmer, Christian, Clemen, Thomas, Hickler, Thomas, Kutsch, Werner Leo, Lenfers, Ulfia A, Martens, Carola, Midgley, Guy F, Mukwashi, Kanisios, Odipo, Victor, Scheiter, Simon, Schmullius, Christiane, Baade, Jussi, du Toit, Justin C, Scholes, Robert J, Smit, Izak P, Stevens, Nicola, Twine, Wayne

  • Authors: Berger, Christian , Bieri, Mari , Bradshaw, Karen L , Brümmer, Christian , Clemen, Thomas , Hickler, Thomas , Kutsch, Werner Leo , Lenfers, Ulfia A , Martens, Carola , Midgley, Guy F , Mukwashi, Kanisios , Odipo, Victor , Scheiter, Simon , Schmullius, Christiane , Baade, Jussi , du Toit, Justin C , Scholes, Robert J , Smit, Izak P , Stevens, Nicola , Twine, Wayne
  • Date: 2019
  • Subjects: To be catalogued
  • Language: English
  • Type: text , article
  • Identifier: http://hdl.handle.net/10962/460589 , vital:75967 , xlink:href="https://doi.org/10.1007/s10584-019-02544-0"
  • Description: Southern Africa is particularly sensitive to climate change, due to both ecological and socio-economic factors, with rural land users among the most vulnerable groups. The provision of information to support climate-relevant decision-making requires an understanding of the projected impacts of change and complex feedbacks within the local ecosystems, as well as local demands on ecosystem services. In this paper, we address the limitation of current approaches for developing management relevant socio-ecological information on the projected impacts of climate change and human activities. We emphasise the need for linking disciplines and approaches by expounding the methodology followed in our two consecutive projects. These projects combine disciplines and levels of measurements from the leaf level (ecophysiology) to the local landscape level (flux measurements) and from the local household level (socio-economic surveys) to the regional level (remote sensing), feeding into a variety of models at multiple scales. Interdisciplinary, multi-scaled, and integrated socio-ecological approaches, as proposed here, are needed to compliment reductionist and linear, scale-specific approaches. Decision support systems are used to integrate and communicate the data and models to the local decision-makers.
  • Full Text:
  • Date Issued: 2019

Segmenting objects with indistinct edges, with application to aerial imagery of vegetation

- James, Katherine M F, Bradshaw, Karen L

  • Authors: James, Katherine M F , Bradshaw, Karen L
  • Date: 2019
  • Subjects: To be catalogued
  • Language: English
  • Type: text , book
  • Identifier: http://hdl.handle.net/10962/460614 , vital:75969 , ISBN 9781450372657 , https://doi.org/10.1145/3351108.3351124
  • Description: Image segmentation mask creation relies on objects having distinct edges. While this may be true for the objects seen in many image segmentation challenges, it is less so when approaching tasks such as segmentation of vegetation in aerial imagery. Such datasets contain indistinct edges, or areas of mixed information at edges, which introduces a level of annotator subjectivity at edge pixels. Existing loss functions apply equal learning ability to both these pixels of low and high annotation confidence. In this paper, we propose a weight map based loss function that takes into account low confidence in the annotation at edges of objects by down-weighting the contribution of these pixels to the overall loss. We examine different weight map designs to find the most optimal one when applied to a dataset of aerial imagery of vegetation, with the task of segmenting a particular genus of shrub from other land cover types. When compared to inverse class frequency weighted binary cross-entropy loss, we found that using weight map-based loss produced a better performing model than binary cross-entropy loss, improving F1 score by 4%.
  • Full Text:
  • Date Issued: 2019
  • Disclaimer
  • Privacy
  • Copyright
  • Contact