Automated Software Testing and Repair Agents
Agent-based software testing, automated program repair, test evolution, flaky-test repair, and large language models for software engineering.
Selected research and engineering projects. My current doctoral work is listed first.
Agent-based software testing, automated program repair, test evolution, flaky-test repair, and large language models for software engineering.
Computer-vision methods for reading hand-drawn logic circuits, comparing YOLO, Faster R-CNN, RetinaNet and Detectron2, with a boundary-tracking method for recovering connections.
Dataset construction and a like-for-like comparison of HOG and LBP feature extraction against deep-learning models.
Estimating vehicle speed from road video using YOLO detection, multi-object tracking and perspective transformation.
COVID-19 classification from chest X-rays, breast-ultrasound segmentation, brain-tumour classification, MRI analysis and melanoma classification.
Human pose-estimation pipelines using MediaPipe and OpenPose to detect and connect body keypoints across video frames.
YOLO-based face detection and real-time facial-expression analysis with Haar cascades, VGG-Face and DeepFace.
A Streamlit dashboard for storing, searching and analysing Telegram messages against a structured database.