Robust relative attributes for human action recognition
Robust relative attributes for human action recognition
复制标题
DOI：
10.1007/s10044-013-0349-3
复制
发表时间：
2015-02
影响因子：
3.9
通讯作者：
Zhong Zhang;Chunheng Wang;Baihua Xiao;Wen Zhou;Shuang Liu
中科院分区：
文献类型：
--
作者：
Zhong Zhang;Chunheng Wang;Baihua Xiao;Wen Zhou;Shuang Liu
高级语义特征是识别人类行为的重要手段。近年来，用于描述相对关系的相对属性作为一种高级语义特征被提出，并表现出良好的性能。然而，训练过程对噪声非常敏感，而且对零射击学习不具有鲁棒性。在本文中，为了克服这些缺点，我们提出了一个使用相对属性进行人类行为识别的鲁棒学习框架。我们同时在损失目标中加入了s型包络和高斯包络。这样，在优化过程中会大大减少异常值的影响，从而提高精度。此外，我们采用高斯混合模型来更好地拟合动作在排名得分空间中的分布。相应的，提出了一种新的转移策略来评估未知类高斯混合模型的参数。我们的方法在KTH、UIUC和HOLLYWOOD2三个具有挑战性的数据集上进行了验证，实验结果表明，我们的方法在人体动作识别的零射击分类和传统识别任务上都取得了比以往方法更好的结果。
High-level semantic feature is important to recognize human action. Recently, relative attributes, which are used to describe relative relationship, have been proposed as one of high-level semantic features and have shown promising performance. However, the training process is very sensitive to noises and moreover it is not robust to zero-shot learning. In this paper, to overcome these drawbacks, we propose a robust learning framework using relative attributes for human action recognition. We simultaneously add Sigmoid and Gaussian envelops into the loss objective. In this way, the influence of outliers will be greatly reduced in the process of optimization, thus improving the accuracy. In addition, we adopt Gaussian Mixture models for better fitting the distribution of actions in rank score space. Correspondingly, a novel transfer strategy is proposed to evaluate the parameters of Gaussian Mixture models for unseen classes. Our method is verified on three challenging datasets (KTH, UIUC and HOLLYWOOD2), and the experimental results demonstrate that our method achieves better results than previous methods in both zero-shot classification and traditional recognition task for human action recognition.