Fourier Shuffle for Semantic Vision
A frequency–spatial upsampling method for medical image segmentation that improves validation performance while reducing memory use.
+1.77% validation · −13.8% GPU memoryBrisbane, Australia
Jevi Waugh
Machine learning researcher working on frequency-domain methods for computer vision. I build the systems I study, and teach the mathematics behind them.
- Machine Learning Research AssistantThe University of QueenslandMar 2026 - Present
- AI Engineer (Part-time)Apex FincapJul 2026 - Present
- Casual AcademicThe University of QueenslandJul 2025 - Present
- Master of Science (Informatics)Technical University of MunichFrom 2027 (intended)
Open to research collaboration
+1.77%validation performanceagainst a U-Net baseline
−13.8%GPU memorywith the Fourier Shuffle model
$35K+sponsorship securedfor UQ Computing Society
3K+extension downloadsfor TraceForest
Selected work
Projects
Research prototypes and developer tools across computer vision, explainable AI, and NLP. Each links to methods, results, and source.
BlackBox Labs: TraceForest
An explainable detector for AI-generated code, combining stylometric AST features, random forests, SHAP, and developer tooling.
94% accuracy · 1.5K+ extension downloads
Vision Transformer Attention Analysis
An experimental pipeline for extracting, visualising, and modifying internal attention distributions in pretrained vision transformers.
FLAN-T5 for Biomedical Lay Summarisation
A comparison of full fine-tuning, LoRA, and evolution strategies for translating specialist biomedical reports into accessible summaries.
>70% ROUGE · 2–3% trainable parameters with LoRALatest talk
COMP3710 Pattern Recognition and Analysis
A guest lecture on frequency-domain autoencoders, covering why spectral representations help a model reconstruct spatial detail and how the architectures are built and evaluated.
15 October 2026
About
Research, engineering, and teaching.
I am a research assistant and Master of Data Science student at The University of Queensland, working on computer vision, deep learning, and the mathematical structure behind model behaviour.
Alongside research I teach capstone computing build, pattern recognition, and introductory software engineering; work part-time as an AI engineer at Apex Fincap; and lead industry partnerships for UQ Computing Society.