BoxPolyp:Boost Generalized Polyp Segmentation Using Extra Coarse Bounding Box Annotations

TL;DR

This paper introduces BoxPolyp, a model that leverages both accurate mask and coarse bounding box annotations to improve polyp segmentation, addressing data scarcity and enhancing model robustness for colorectal cancer diagnosis.

Contribution

We propose a boosted segmentation framework utilizing a fusion filter sampling module and an image consistency loss, which can be integrated into any backbone to improve polyp segmentation performance.

Findings

Outperforms state-of-the-art methods on five benchmarks

Significant performance gains with coarse box annotations

Enhanced robustness and generalization of the segmentation model

Abstract

Accurate polyp segmentation is of great importance for colorectal cancer diagnosis and treatment. However, due to the high cost of producing accurate mask annotations, existing polyp segmentation methods suffer from severe data shortage and impaired model generalization. Reversely, coarse polyp bounding box annotations are more accessible. Thus, in this paper, we propose a boosted BoxPolyp model to make full use of both accurate mask and extra coarse box annotations. In practice, box annotations are applied to alleviate the over-fitting issue of previous polyp segmentation models, which generate fine-grained polyp area through the iterative boosted segmentation model. To achieve this goal, a fusion filter sampling (FFS) module is firstly proposed to generate pixel-wise pseudo labels from box annotations with less noise, leading to significant performance improvements. Besides, considering the appearance consistency of the same polyp, an image consistency (IC) loss is designed. Such IC loss explicitly narrows the distance between features extracted by two different networks, which improves the robustness of the model. Note that our BoxPolyp is a plug-and-play model, which can be merged into any appealing backbone. Quantitative and qualitative experimental results on five challenging benchmarks confirm that our proposed model outperforms previous state-of-the-art methods by a large margin.