Bounding Box Priors for Cell Detection with Point Annotations

TL;DR

This paper introduces a novel weakly semi-supervised cell detection method using bounding box priors and stochastic boxes, outperforming existing approaches especially with limited box-annotated data.

Contribution

It proposes a one-stage training approach combining point and box annotations with mean-IOU stochastic boxes, improving detection accuracy in data-scarce scenarios.

Findings

Outperforms two-stage methods by 5.56 mAP with only 5% box-labeled images.

Uses class-specific prior distributions for bounding boxes to enhance detection.

Provides insights on the minimal amount of box annotations needed for effective training.

Abstract

The size of an individual cell type, such as a red blood cell, does not vary much among humans. We use this knowledge as a prior for classifying and detecting cells in images with only a few ground truth bounding box annotations, while most of the cells are annotated with points. This setting leads to weakly semi-supervised learning. We propose replacing points with either stochastic (ST) boxes or bounding box predictions during the training process. The proposed "mean-IOU" ST box maximizes the overlap with all the boxes belonging to the sample space with a class-specific approximated prior probability distribution of bounding boxes. Our method trains with both box- and point-labelled images in conjunction, unlike the existing methods, which train first with box- and then point-labelled images. In the most challenging setting, when only 5% images are box-labelled, quantitative experiments on a urine dataset show that our one-stage method outperforms two-stage methods by 5.56 mAP. Furthermore, we suggest an approach that partially answers "how many box-labelled annotations are necessary?" before training a machine learning model.