A Semi-Supervised Data Augmentation Approach using 3D Graphical Engines-猫眼课题宝

DOI：

10.1007/978-3-030-11012-3_31

发表时间：

2018-08

期刊：

影响因子：

通讯作者：

中科院分区：

其他

文献类型：

作者：

Shuangjun Liu;S. Ostadabbas

深度学习方法由于其准确性和灵活性而在广泛的领域中迅速被采用，但需要大量的标记训练数据集。这对于具有有限、昂贵或私有数据（即小数据）的应用提出了一个基本问题，例如可以高度个性化的人类姿势和行为估计/跟踪。在本文中，我们提出了一种半监督的数据增强方法，可以合成大规模的标记训练数据集使用3D图形引擎的基础上，物理上有效的低维姿态描述符。为了评估我们的合成数据集在训练基于深度学习的模型方面的性能，我们基于我们提出的增强方法，使用仅7个人的3D扫描生成了一个大型合成人体姿势数据集，称为ScanAva。然后，使用我们的ScanAva数据集从头开始训练最先进的人体姿势估计深度学习模型，并在PCK 0时实现91.2%的姿势估计准确度。在对合成图像应用有效的域自适应之后，其姿态估计精度与在来自真实的人类的大规模姿态数据（诸如MPII数据集）上训练的相同模型相当，并且远高于在其他合成人类数据集（诸如SURREAL）上训练的模型。

Deep learning approaches have been rapidly adopted across a wide range of fields because of their accuracy and flexibility, but require large labeled training datasets. This presents a fundamental problem for applications with limited, expensive, or private data (ie small data), such as human pose and behavior estimation/tracking which could be highly personalized. In this paper, we present a semi-supervised data augmentation approach that can synthesize large scale labeled training datasets using 3D graphical engines based on a physically-valid low dimensional pose descriptor. To evaluate the performance of our synthesized datasets in training deep learning-based models, we generated a large synthetic human pose dataset, called ScanAva using 3D scans of only 7 individuals based on our proposed augmentation approach. A state-of-the-art human pose estimation deep learning model then was trained from scratch using our ScanAva dataset and could achieve the pose estimation accuracy of 91.2% at PCK0. 5 criteria after applying an efficient domain adaptation on the synthetic images, in which its pose estimation accuracy was comparable to the same model trained on large scale pose data from real humans such as MPII dataset and much higher than the model trained on other synthetic human dataset such as SURREAL.

A Semi-Supervised Data Augmentation Approach using 3D Graphical Engines

文献摘要

相似文献