PolarMask: Single Shot Instance Segmentation with Polar Representation

TL;DR

PolarMask introduces a simple, fully convolutional, single shot instance segmentation method using polar coordinates, achieving competitive accuracy on COCO without complex components.

Contribution

It presents a novel polar coordinate-based framework for instance segmentation that is anchor-box free, easy to embed, and improves training simplicity and performance.

Findings

Achieves 32.9% mask mAP on COCO with single-model, single-scale training/testing.

Formulates segmentation as center classification and dense distance regression in polar coordinates.

Demonstrates a flexible, competitive baseline for single shot instance segmentation.

Abstract

In this paper, we introduce an anchor-box free and single shot instance segmentation method, which is conceptually simple, fully convolutional and can be used as a mask prediction module for instance segmentation, by easily embedding it into most off-the-shelf detection methods. Our method, termed PolarMask, formulates the instance segmentation problem as instance center classification and dense distance regression in a polar coordinate. Moreover, we propose two effective approaches to deal with sampling high-quality center examples and optimization for dense distance regression, respectively, which can significantly improve the performance and simplify the training process. Without any bells and whistles, PolarMask achieves 32.9% in mask mAP with single-model and single-scale training/testing on challenging COCO dataset. For the first time, we demonstrate a much simpler and flexible instance segmentation framework achieving competitive accuracy. We hope that the proposed PolarMask framework can serve as a fundamental and strong baseline for single shot instance segmentation tasks. Code is available at: github.com/xieenze/PolarMask.