Research

Papers & whitepapers.

The longer, more formal version of the work. Recent papers are on verifying AI-written software; the earlier ones come from my computer vision research in Malaysia. Each can be read in the browser or downloaded as a PDF.

Also on Google Scholar and ORCID.

01
Preprint · Zenodo

Cleared the Bar, or Lowered It? An Assurance Lattice for Verifying AI-Generated Software Against Intent→

Coding agents can turn every check green by quietly weakening the requirement. This paper sets out a theory of assurance that tells a bar that was cleared from a bar that was lowered.

AI-generated codeSoftware verificationRequirements traceabilityLLM agentsLattice
Oct 2026
PDF
02
Paper · Jurnal Penyelidikan Berimpak Tinggi MARA

Development of Real-Time Vehicle's Plate Number Recognition System Using Deep Learning Approach for Surveillance System in Malaysia→

A mobile system that detects and reads Malaysian license plates in real time with deep learning, built for vehicle surveillance in moving, poorly lit, complex scenes.

License plate recognitionDeep learningIntelligent transportationMobile
Nov 2022
PDF
03
Paper · Malaysian Journal of Computer Science

License Plate Recognition Techniques: Comparative Study→

A survey of how license plate recognition systems work, comparing traditional computer vision with deep learning on accuracy, performance, strengths and weaknesses.

License plate recognitionComputer visionDeep learning
Dec 2021
PDF