Back to portfolioAI Products
Parasite Egg Detection
YOLOv11m image detection for microscope smear images
AI ProductsPA
Parasite Egg Detection
PythonYOLOv11mPyTorch
The Problem
Manual identification of parasite eggs in microscope smear images is slow, requires expert pathologists, and is error-prone -- limiting access to parasitology screening in resource-constrained settings.
The Solution
Built a fine-tuned YOLOv11m detector with paper-inspired hyperparameters (640px imgsz, AdamW, lr0=0.001667) and aggressive augmentation (rotation, shear, flips, HSV) tuned for thick-smear specimens. Submitted to Super AI Engineer Season 6.
Tech Stack
PythonYOLOv11mPyTorchUltralyticsOpenCV
Key Features
Fine-tuned YOLOv11m detector
Paper-inspired training config
Augmentation tuned for thick-smear specimens
Submission pipeline with detect-then-crop workflow
Built for Super AI Engineer Season 6