logo
科言猫学术研究的AI总结
language
手机端手机端
  • Phone Feature微信小程序
简体中文简体中文
notice消息通知
博客博客
delete
arrow
返回

PSINet: Parallel Space Interaction Network for Human Pose Estimation

delete2026-01-12
delete0
PRE
AI
D
Dong Wang
Y
Youcheng Cai
唐
唐益明 (Yiming Tang)
X
Xiaoping Liu
DOI：10.1109/TETCI.2025.3647687delete
delete原文链接
delete原文求助
delete分享
delete收藏
摘要

摘要

En 中文
人体姿态估计对单幅图像中特征表示的空间分辨率敏感。虽然多尺度特征常被用于改进关键点定位，但跨多个统一尺寸空间的并行空间学习的有效性仍需探索。为解决此问题，我们提出并行空间交互网络（PSINet），将输入特征分割为多个相同的并行空间，以促进人体姿态估计的高效空间特征学习。基于多个空间，我们通过提出分裂头Transformer（SHT）来改进原始视觉Transformer，以增强并行空间内的全局交互。此外，我们引入并行空间融合模块以通过获取多个并行空间之间的关系来增强PSINet的能力。大量实验表明，经过数据增强后，PSINet达到最先进精度，在单块GTX2080Ti上以180 FPS运行，在精度和效率之间提供卓越的权衡。
Keyword:
Parallel space interaction
transformer
human pose estimation
feature interaction

期刊

I
IEEE Transactions on Emerging Topics in Computational Intelligence
IF:
6.5
论文数:
1.4K
被引数:
4.5K

机构

H
hefei university of technology
学者数:
2.5W
论文数: 1.7W
被引数: 35
A
anhui jianzhu university
学者数:
1.3K
论文数: 474
被引数: 0
U
University of Science and Technology of China
学者数:
1.7W
论文数: 6.0K
被引数: 11.3W
学者 查看更多机构
引用论文

引用论文

TransPose: Keypoint Localization via Transformer
err2021 IEEE/CVF International Conference on Computer Vision (ICCV)
IF0
err2021-10-01
err0
errOAAI
errSen Yang; Zhibin Quan; Mu Nie; Wankou Yang
err分享
err收藏
Scene-Aware Egocentric 3D Human Pose Estimation
err2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)
IF0
err2023-06-01
err0
errOAAI
errJian Wang; Diogo Luvizon; Weipeng Xu; Lingjie Liu; Kripasindhu Sarkar; Christian Theobalt
err分享
err收藏
FASTCC: A lightweight human pose detection method leveraging SimCCFASTCC: 一种利用SimCC的轻量级人体姿态检测方法
errNEUROCOMPUTING
IF6.5
err2025-04-01
err0
PREAI
errLi, Yi; Zhang, Jialin; Wang, Yongtao; Quan, Dou; Yan, Yabo; Yang, Qinghai
err分享
err收藏
UIE-Convformer: Underwater Image Enhancement Based on Convolution and Feature Fusion TransformerUie-convformer: 基于卷积和特征融合变换器的水下图像增强
errIEEE TRANSACTIONS ON EMERGING TOPICS IN COMPUTATIONAL INTELLIGENCE
IF6.5
err2024-04-01
err10
PREAI
errWang, Biao; Xu, Haiyong; Jiang, Gangyi; Yu, Mei; Ren, Tingdi; Luo, Ting; Zhu, Zhongjie
err分享
err收藏
2D Human Pose Estimation: New Benchmark and State of the Art Analysis
err2014 IEEE Conference on Computer Vision and Pattern Recognition
IF0
err2014-06-01
err0
PREAI
errMykhaylo Andriluka; Leonid Pishchulin; Peter Gehler; Bernt Schiele
err分享
err收藏
SG-Former: Self-guided Transformer with Evolving Token Reallocation
err2023 IEEE/CVF International Conference on Computer Vision (ICCV)
IF0
err2023-10-01
err0
errOAAI
errSucheng Ren; Xingyi Yang; Songhua Liu; Xinchao Wang
err分享
err收藏
The Devil Is in the Details: Delving Into Unbiased Data Processing for Human Pose Estimation
err2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)
IF0
err2020-06-01
err0
errOAAI
errJunjie Huang; Zheng Zhu; Feng Guo; Guan Huang
err分享
err收藏
Lite Pose: Efficient Architecture Design for 2D Human Pose Estimation
err2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)
IF0
err2022-06-01
err0
errOAAI
errYihan Wang; Muyang Li; Han Cai; Weiming Chen; Song Han
err分享
err收藏
Attention Is All You Need注意是你所需要的
err
IF0
err
err0
errOAAI
err
err分享
err收藏
MTPose: Human Pose Estimation with High-Resolution Multi-scale Transformers
errNeural Processing Letters
IF0
err2022-03-29
err0
PREAI
errRui Wang; Fudi Geng; Xiangyang Wang
err分享
err收藏
学者 查看更多内容