Quality-Aware Network for Human Parsing

doi:10.1109/TMM.2022.3217413

CASIA OpenIR > 紫东太初大模型研究中心

	Quality-Aware Network for Human Parsing
	Yang, Lu 1; Song, Qing 1; Wang, Zhihui 1; Liu, Zhiwei2 ; Xu, Songcen 3; Li, Zhihao 3
发表期刊	IEEE TRANSACTIONS ON MULTIMEDIA
ISSN	1520-9210
	2023
卷号	25 页码:7128-7138
通讯作者	Song, Qing(priv@bupt.edu.cn)
摘要	How to estimate the quality of the network output is an important issue, and currently there is no effective solution in the field of human parsing. To solve this problem, this work proposes a statistical method based on the output probability map to calculate the pixel classification quality, which is called pixel score. In addition, the Quality-Aware Module (QAM) is proposed to fuse the different quality information, the purpose of which is to estimate the quality of human parsing results. We combine QAM with a concise and effective network design to propose Quality-Aware Network (QANet) for human parsing. Benefiting from the superiority of QAM and QANet, we achieve the best performance on three multiple and one single human parsing benchmarks, including CIHP, MHP-v2, Pascal-Person-Part, ATR and LIP. Without increasing the training and inference time, QAM improves the AP$<^>\text{r}$ criterion by more than 10 points in the multiple human parsing task. QAM can be extended to other tasks with good quality estimation, e.g instance segmentation. Specifically, QAM improves Mask R-CNN by similar to% mAP on COCO and LVISv1.0 datasets. Based on the proposed QAM and QANet, our overall system wins 1st place in CVPR2021 L2ID High-resolution Human Parsing (HRHP) Challenge, and 2nd in CVPR2021 PIC Short-video Face Parsing (SFP) Challenge. Code and models are available at https://github.com/soeaver/QANet.
关键词	Computer vision image segmentation multi-media computing
DOI	10.1109/TMM.2022.3217413
关键词[WOS]	POSE
收录类别	SCI
语种	英语
资助项目	China Postdoctoral Science Foundation[2022M710467] ; National Key Research and Development Program of China[2021YFF0500900]
项目资助者	China Postdoctoral Science Foundation ; National Key Research and Development Program of China
WOS研究方向	Computer Science ; Telecommunications
WOS类目	Computer Science, Information Systems ; Computer Science, Software Engineering ; Telecommunications
WOS记录号	WOS:001102654000030
出版者	IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
引用统计	被引频次：4[WOS] [WOS记录] [WOS相关记录]
文献类型	期刊论文
条目标识符	http://ir.ia.ac.cn/handle/173211/55163
专题	紫东太初大模型研究中心
通讯作者	Song, Qing
作者单位	1.Beijing Univ Posts & Telecommun, Beijing 100088, Peoples R China 2.Chinese Acad Sci, Inst Automat, Beijing 100190, Peoples R China 3.Huawei Technol, Noahs Ark Lab, Shenzhen 518129, Peoples R China
推荐引用方式 GB/T 7714	Yang, Lu,Song, Qing,Wang, Zhihui,et al. Quality-Aware Network for Human Parsing[J]. IEEE TRANSACTIONS ON MULTIMEDIA,2023,25:7128-7138.
APA	Yang, Lu,Song, Qing,Wang, Zhihui,Liu, Zhiwei,Xu, Songcen,&Li, Zhihao.(2023).Quality-Aware Network for Human Parsing.IEEE TRANSACTIONS ON MULTIMEDIA,25,7128-7138.
MLA	Yang, Lu,et al."Quality-Aware Network for Human Parsing".IEEE TRANSACTIONS ON MULTIMEDIA 25(2023):7128-7138.