MSP : Refine Boundary Segmentation via Multiscale Superpixel

TL;DR

This paper introduces MSP, a multiscale superpixel-based message passing module that enhances boundary sharpness in semantic segmentation, applicable to various networks and datasets, improving boundary quality without extra parameters.

Contribution

The paper presents a novel multiscale superpixel module that refines boundary segmentation by guiding message passing, compatible with existing networks and datasets.

Findings

Improves boundary sharpness in segmentation results.

Effective across multiple baseline models and datasets.

No additional parameters needed for integration.

Abstract

In this paper, we propose a simple but effective message passing method to improve the boundary quality for the semantic segmentation result. Inspired by the generated sharp edges of superpixel blocks, we employ superpixel to guide the information passing within feature map. Simultaneously, the sharp boundaries of the blocks also restrict the message passing scope. Specifically, we average features that the superpixel block covers within feature map, and add the result back to each feature vector. Further, to obtain sharper edges and farther spatial dependence, we develop a multiscale superpixel module (MSP) by a cascade of different scales superpixel blocks. Our method can be served as a plug-and-play module and easily inserted into any segmentation network without introducing new parameters. Extensive experiments are conducted on three strong baselines, namely PSPNet, DeeplabV3, and DeepLabV3+, and four challenging scene parsing datasets including ADE20K, Cityscapes, PASCAL VOC, and PASCAL Context. The experimental results verify its effectiveness and generalizability.