Multimodal Activity Recognition
For my individual MSc dissertation, I prepared pose and wearable sensor data, implemented CNN-LSTM fusion in PyTorch and evaluated it against single-modality models.
READ CASE STUDY
I work across machine learning, deep learning and AI optimisation, from preparing data and developing models to evaluating results. I also build Power BI dashboards for business reporting.
EXPLORE WORKGET IN TOUCHMy projects cover multimodal learning, computer vision, robotics and optimisation. Alongside this research, I have built Power BI dashboards for business reporting.
For my individual MSc dissertation, I prepared pose and wearable sensor data, implemented CNN-LSTM fusion in PyTorch and evaluated it against single-modality models.
READ CASE STUDYI gathered and labelled image data, trained YOLOv4-tiny and wrote Python/OpenCV detection code for a team-built battery-sorting robot.
READ CASE STUDY
I developed an NSGA-II-inspired scheduling optimiser and used an external archive and visualisations to explore worker usage, workload balance and idle time.
READ CASE STUDYIn my data analyst role, I developed Power BI dashboards with DAX and combined SQL and Excel data for reporting, including data cleaning and transformation.
I prepared MediaPipe landmark data and developed MLP, XGBoost and ensemble classifiers for yoga-pose recognition within a group stereo-vision project.
Computer Vision / Group ProjectGroup coursework analysing 655 customer survey records in R, with regression, variance testing and ggplot2. The report also explores business-news text and sentiment.
Data Analysis / Group ProjectI built a YUV segmentation and contour pipeline for droplet counting, and compared baseline and deeper regularised CNNs on CIFAR-10.
Deep Learning / Individual ProjectCoursework covering Q-learning, boosting, mushroom classification and a comparison of KMeans and Gaussian mixture clustering on white-wine data.
Machine Learning / Individual AssessmentI reviewed and compared GANs, VAEs, diffusion, autoregressive and normalizing-flow models, examining computational trade-offs and ethical considerations.
Literature Review / Individual AssessmentAcademic work spans individual research and team projects. Case studies distinguish reported results from deployment goals and identify the limits of the available evidence.
I'm Doniel O. Amosu, an Artificial Intelligence postgraduate from the University of Plymouth with a Computer Science background and professional experience in data analytics, business intelligence and front-end development.
My work spans machine learning, deep learning and optimisation, including activity recognition, computer vision, classification, clustering and multi-objective scheduling. I also build Power BI dashboards, prepare data with SQL and develop web interfaces.
My coursework also includes Q-learning and a research review of generative models, including GANs, VAEs and diffusion. These studies complement my practical modelling and evaluation work.
My frontline experience in UK healthcare has also shaped my interest in how data and intelligent systems could support better services and decisions. Health Tech is an area I aim to grow into, alongside opportunities in data and AI.
SQL, Power BI, DAX, Excel, Python, R, pandas, NumPy, ggplot2, statistical testing and reporting.
Python, scikit-learn, TensorFlow, PyTorch, OpenCV, XGBoost, CNNs, LSTMs, classification, clustering, model evaluation and NSGA-II optimisation. Coursework includes Q-learning and generative model research.
HTML, CSS, JavaScript, Git/GitHub, SQL databases and Microsoft Azure.
Data cleaning, visualisation, technical documentation, collaboration and coordinating academic project deliverables.
Developed Power BI dashboards with DAX, combined SQL and Excel data for reporting, and investigated data inconsistencies to improve reporting reliability.
Built responsive client-facing web applications, supported SQL and database tasks, and used Azure and Power BI in development and reporting workflows.
Frontline healthcare work alongside my transition into technology, informing my interest in practical Health Tech applications.
Working in UK healthcare has given me a practical perspective on the people and processes behind care. I'm interested in exploring how responsible AI, analytics and intelligent systems might support healthcare operations and improve the use of information. I’m particularly interested in healthcare operations, useful information systems and responsible evaluation of AI in practice.
I'm open to opportunities in data analysis, business intelligence and AI-related roles, as well as collaborations where technical research can be put to practical use.
EMAIL MEDOWNLOAD CV (PDF)LINKEDINGITHUB