PolarMask: Single Shot Instance Segmentation With Polar Representation

By Enze Xie, Pei Sun, Xiaoge Song, Wenhai Wang, Xuebo Liu, Ding Liang, Chunhua Shen, Ping Luo. 2019

627 citations recorded by the source corpus.

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 by easily embedding it into most off-the-shelf detection methods. Our method, termed PolarMask, formulates the instance segmentation problem as predicting contour of instance through 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 the challenging COCO dataset. For the first time, we show that the complexity of instance segmentation, in terms of both design and computation complexity, can be the same as bounding box object detection and this much simpler and flexible instance segmentation framework can achieve competitive accuracy. We hope that the proposed PolarMask framework can serve as a fundamental and strong baseline for single shot instance segmentation task.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computer Vision and Pattern Recognition

Citations in the corpus

iShape: A First Step Towards Irregular Shape Instance Segmentation (2021, neutral): “Instead, works like [15] tackle the problem within a single-stage manner.”
Pyramid Vision Transformer: A Versatile Backbone for Dense Prediction without Convolutions (2021, background): “, SSD [42], RetinaNet [38], FCOS [61], GFL [36, 34], PolarMask [70] and OneNet [54]) and multi-stage detectors (Faster R-CNN [48], Mask R-CNN [20], Cascade R-CNN [4] and Sparse R-CNN [56]).”
Image Segmentation Using Deep Learning: A Survey (2020, methodology): “Other instance segmentation models have been developed based on R-CNN, such as those developed for mask proposals, including R-FCN [68], DeepMask [69], PolarMask [70], boundary-aware instance segmentation [71], and CenterMask [72].”
SOLOv2: Dynamic and Fast Instance Segmentation (2020, methodology): “In particular, recent methods of [4, 38, 35] build their methods on the anchor-free object detectors [31], showing promising performance.”
MapTR: Structured Modeling and Learning for Online Vectorized HD Map Construction (2022, background): “Another line of work related to MapTR is contour-based 2D instance segmentation (Zhu et al., 2022; Xie et al., 2020; Xu et al., 2019; Liu et al., 2021c).”
Interactive Object Segmentation With Inside-Outside Guidance (2020, background): “Over the past few years, we have witnessed a revolutionary advancement in semantic [44, 40, 68, 69, 10, 11, 12, 30, 15, 31] and instance segmentation [25, 36, 8, 60, 13, 65, 34, 4, 9, 43, 29] for different domains, such as general scenes [20, 41, 70], autonomous driving [17, 48, 21], aerial imagery ”
IoU-aware Single-stage Object Detector for Accurate Localization (2019, methodology): “PolarMask [23] modifies FCOS to realize the instance segmentation.”
YOLACT++ Better Real-Time Instance Segmentation (2019, background): “, [3], [8], [9]), but they require significant amounts of post-processing after localization, and thus are still far from real-time.”
BoxInst: High-Performance Instance Segmentation with Box Annotations (2020, methodology): “BoxInst, without using any mask annotations, performs even better than some recent fully-supervised methods such as PolarMask [31] and YOLACT [3] , with the same backbones as well as similar training and testing settings (32.5% with R-101 1 × vs. PolarMask 32.1% R-101 2 × and YOLACT-700 31.2% R-101 ”
Learning What Not to Segment: A New Perspective on Few-Shot Segmentation (2022, background): “However, collecting sufficient labeled data is notoriously time-consuming and labor-intensive, especially for dense prediction tasks, such as instance segmentation [2,3,15,21,59] and semantic segmentation [1,25,35,40,45].”
Interactive Object Segmentation With Inside-Outside Guidance (2022, background): “OVER the past few years, we have witnessed a revolutionary advancement in semantic [9], [10], [11], [12], [13], [14], [15], [16], [17], [18] and instance segmentation [19], [20], [21], [22], [23], [24], [25], [26], [27], [28], [29] for different domains, such as general scenes [3], [30], [31], auton”
DiscoBox: Weakly Supervised Instance Segmentation and Semantic Correspondence from Box Supervision (2021, methodology): “Our best model achieves 37.9% AP on COCO test-dev, surpassing competitive supervised methods such as YOLACT++ [50] (34.6% AP) and Mask R-CNN [51] (37.1% AP).”