首页> 外文期刊>Pattern Recognition: The Journal of the Pattern Recognition Society >A CNN-based 3D human pose estimation based on projection of depth and ridge data
【24h】

A CNN-based 3D human pose estimation based on projection of depth and ridge data

机译:基于CNN的3D人体姿势估计基于深度和脊数据的投影

获取原文
获取原文并翻译 | 示例
           

摘要

We propose a method that use a convolutional neural network (CNN) to estimate human pose by analyzing the projection of the depth and ridge data, which represent local maxima in a distance transform map. To fully utilize the 3D information of depth points, we propose a method to project the depth and ridge data on various directions. The proposed projection method can reduce the 3D information loss, the ridge data can avoid joint drift, and the CNN increases localization accuracy. The proposed method proceeds as follows. (1) We use depth data to segment the human from the background and extract ridge data from human silhouettes. (2) We project the depth and ridge data onto XY, XZ, and ZY planes. (3) ResNet-101 accepts six projected images and use 1 x 1 convolution layers to generate 2D heatmaps and offsets. (4) We generate 2D keypoints per plane by using the soft-argmax operation. (5) We obtain 3D joint positions by using the fully-connected layers. In experiments on the SMMC-10, EVAL, and ITOP datasets, the proposed method achieved the state-of-the-art pose estimation accuracies. The proposed method can eliminate the 3D information loss and drift of joint positions that can occur during estimation of human pose. Keywords: 3D Human pose estimation 3D Point projection Ridge data (C) 2020 Elsevier Ltd. All rights reserved.
机译:我们提出了一种使用卷积神经网络(CNN)来估计人类姿势的方法,通过分析深度和脊数据的投影,这在距离变换图中表示局部最大值。为了充分利用深度点的3D信息,我们提出了一种在各种方向上投射深度和脊数据的方法。所提出的投影方法可以减少3D信息丢失,脊数据可以避免关节漂移,并且CNN增加了本地化精度。所提出的方法如下所述。 (1)我们使用深度数据将人类从背景中分段并从人剪影中提取脊数据。 (2)我们将深度和脊数据投影到XY,XZ和ZY平面上。 (3)Resnet-101接受六个投影图像,并使用1 x 1卷积图层来生成2D热插拔和偏移。 (4)通过使用软氩操作,我们每平面生成2D关键点。 (5)我们通过使用完全连接的层获得3D接头位置。在SMMC-10,EVAL和ITAP数据集的实验中,所提出的方法实现了最先进的姿态估计精度。该方法可以消除在人类姿势估计期间可以发生的3D信息丢失和漂移。关键词:3D人类姿态估计3D点投影脊数据(c)2020 elsevier有限公司保留所有权利。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号