Description: Developed an image processing pipeline to detect AgNORs in histopathology images, including custom image augmentations, complete model for classification, and object detection mmodel using PyTorch. Achievements: Implemented image classification with an F1-score performance measure. Built an object detection model with a complete pipeline from initialization to performance reporting via AP. Developed a two-stage pipeline combining cell detection and classification for computing an average AgNOR-score.
(PyTorch, Custom Dataset Handling, Albumentations)