您的位置: 首页 > 文章 > iccv 2017-人体姿态估计 Learning Feature Pyramids for Human Pose Estimation

iccv 2017-人体姿态估计 Learning Feature Pyramids for Human Pose Estimation

分类: 文章 • 2023-01-07 00:31:43

读书笔记来自于https://mp.weixin.qq.com/s/FQEK8slP7dWosiTlElsnjA

Learning Feature Pyramids for Human Pose Estimation 是发表在 ICCV 2017 的一篇有关人体姿态估计的论文，提出利用特征金字塔来进行人体姿势预测。作者是 Wei Yang，香港中文大学博士生。

iccv 2017-人体姿态估计 Learning Feature Pyramids for Human Pose Estimation

iccv 2017-人体姿态估计 Learning Feature Pyramids for Human Pose Estimation

iccv 2017-人体姿态估计 Learning Feature Pyramids for Human Pose Estimation

iccv 2017-人体姿态估计 Learning Feature Pyramids for Human Pose Estimation

iccv 2017-人体姿态估计 Learning Feature Pyramids for Human Pose Estimation

改进思路：

1.inception的多分支；2，ResNeXt中的分组卷积；3，dilated convolution

论文中的提出的模块：Pyramid Residual Modules (PRMs)

iccv 2017-人体姿态估计 Learning Feature Pyramids for Human Pose Estimation

iccv 2017-人体姿态估计 Learning Feature Pyramids for Human Pose Estimation

iccv 2017-人体姿态估计 Learning Feature Pyramids for Human Pose Estimation

iccv 2017-人体姿态估计 Learning Feature Pyramids for Human Pose Estimation

iccv 2017-人体姿态估计 Learning Feature Pyramids for Human Pose Estimation

iccv 2017-人体姿态估计 Learning Feature Pyramids for Human Pose Estimation