论文标题
ICCV 2019的联合可可和马皮拉里研讨会：可可实例分割挑战赛道
Joint COCO and Mapillary Workshop at ICCV 2019: COCO Instance Segmentation Challenge Track
论文作者
论文摘要
在本报告中，我们介绍了我们的对象检测/实例分割系统MEGDETV2，该系统以两次通用方式起作用，首先是检测实例然后获得分割的实例。我们的基线检测器主要建立在新设计的RPN上，称为RPN ++。在2019年COCO-2019检测/实例分割测试-DEV数据集上，我们的系统达到了61.0/53.1地图，该地图分别超过了我们2018年的获胜结果，分别为5.0/4.2。我们在2019年和2020年可可挑战赛中取得了最佳成绩。
In this report, we present our object detection/instance segmentation system, MegDetV2, which works in a two-pass fashion, first to detect instances then to obtain segmentation. Our baseline detector is mainly built on a new designed RPN, called RPN++. On the COCO-2019 detection/instance-segmentation test-dev dataset, our system achieves 61.0/53.1 mAP, which surpassed our 2018 winning results by 5.0/4.2 respectively. We achieve the best results in COCO Challenge 2019 and 2020.
