Key Contributions:

• Designed and developed a Streamlit interface with sequential, step-by-step menus guiding users through the full pipeline.
• Implemented video ingestion, automated object detection, and tracking of unique litter items.
• Integrated a human-in-the-loop review stage for validating and correcting detections.
• Generated annotated video outputs to support quality assurance.
• Added AI-driven analytics to enhance insights and reporting.
• Enabled iterative model fine-tuning to continuously improve detection performance.
• Deployed optimised models to edge devices for smooth, real-time monitoring.
• Authored comprehensive documentation outlining two workflows: object detection and model improvement.

Key Contributions:

• Built a deep learning system to estimate floating litter volumes in waterways.
• Identified hotspot regions to support environmental monitoring and ecosystem management.
• Applied advanced CV models (YOLO, Deeplab, Transformers) for multi-task detection.

Key Contributions:

• Developed a standalone, offline web application for running ML detection/classification models.
• Designed interactive dashboards for image analytics and summarised insights.
• Integrated Folium to visualise spatial patterns and geospatial detection outputs through interactive maps and heatmaps.

Key Contributions:

• Analysed large-scale employee review data during the COVID period.
• Identified shifting trends in workplace expectations as part of a research collaboration with University of Melbourne.
• Conducted scraping, pre-processing, feature engineering, and sentiment modelling.

Key Contributions:

• Contributed to an AI funded pilot project to identify individual koalas at road-crossing points.
• Supported model development to aid conservation and road safety strategies in South East Queensland.
• Performed data labeling, preprocessing, and deep learning experiments for facial recognition.

Key Contributions:

• Built and evaluated deep learning models to detect road safety attributes from annual road-survey videos provided by DTMR.
• Integrated object detection and semantic segmentation methods.
• Implemented distance estimation from monocular video data and supported model deployment.

Key Contributions:

• Developed an automated crack detection pipeline using images of asphalt roads and bridges.
• Applied augmentation strategies to overcome limited dataset availability.
• Implemented foreground-background classification and semantic segmentation methods.

Key Contributions:

• Designed algorithms to minimise respiratory motion artifacts in PET/CT images.
• Enhanced image reconstruction accuracy to support improved diagnostic confidence.

Key Contributions:

• Developed a dimension-reduced EMD algorithm for efficient image matching.
• Applied the technique for classifying handwritten samples.
• Designed feature extraction and mathematical optimisation routines.

Key Contributions:

• Designed and implemented enterprise software packages using Borland Delphi and SQL Server.
• Provided technical support, troubleshooting, debugging, and software maintenance.
• Managed installation, configuration, and deployment of solutions at client organisations.

Key Contributions:

• Developed software components using C#, ASP.NET, and SQL Server.
• Conducted debugging, testing, and diagnostics of deployed applications.
• Delivered maintainable and standards-compliant software artefacts.