arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2008.00460cs.CV
收藏

用于实例分割的联合对象轮廓点和语义

Joint Object Contour Points and Semantics for Instance Segmentation

Northeastern University(东北大学);Beijing Information Science and Technology University(北京信息科技大学);Science and Technology on Space Physics Laboratory(空间物理重点实验室);The University of Auckland(奥克兰大学)

Wenchao Zhang, Chong Fu, Mai Zhu

摘要:

目标轮廓的属性对于实例分割任务具有重要意义。然而,目前流行的大部分深度神经网络对物体边缘信息关注不多。受制作实例分割数据集时人类注释过程的启发,本文提出了Mask Point R-CNN,旨在促进神经网络对对象边界的关注。具体来说,我们创新性地将原始的人体关键点检测任务扩展到任何物体的轮廓点检测。基于这种类比,我们为Mask R-CNN提出了一个轮廓点检测辅助任务,它可以通过有效地使用特征融合策略和多任务联合训练来增强不同任务之间的梯度流。因此,模型将对对象的边缘更敏感,并且可以捕获更多的几何特征。定量地,实验结果表明,我们的方法在Cityscapes数据集上比vanilla Mask R-CNN高3.8%,在COCO数据集上高0.8%。

英文摘要:

The attributes of object contours has great significance for instance segmentation task. However, most of the current popular deep neural networks do not pay much attention to the object edge information. Inspired by the human annotation process when making instance segmentation datasets, in this paper, we propose Mask Point R-CNN aiming at promoting the neural network's attention to the object boundary. Specifically, we innovatively extend the original human keypoint detection task to the contour point detection of any object. Based on this analogy, we present an contour point detection auxiliary task to Mask R-CNN, which can boost the gradient flow between different tasks by effectively using feature fusion strategies and multi-task joint training. As a consequence, the model will be more sensitive to the edges of the object and can capture more geometric features. Quantitatively, the experimental results show that our approach outperforms vanilla Mask R-CNN by 3.8\% on Cityscapes dataset and 0.8\% on COCO dataset.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑