论文标题
学习类动作识别的类正规功能
Learning Class Regularized Features for Action Recognition
论文作者
论文摘要
训练深度卷积神经网络（CNN）是基于在随后的激活中使用多个内核和非线性的概念来提取有用的特征。内核用作一般特征提取器，而没有与目标类别的特定对应关系。结果，提取的功能与特定类别不符。相似类之间的细微差异与不同类别之间的巨大差异相同。为了克服CNN中核的类别使用，我们引入了一种名为“类正则化的新方法”，该方法对基于类的层激活进行了基于类的正则化。我们证明，这不仅可以改善训练期间的功能搜索，而且还允许在功能提取过程的每个阶段中明确分配每个课程的功能。我们表明，在最先进的CNN体系结构中使用类正规化块进行动作识别，可导致动力学，UCF-101和HMDB-51数据集的系统改善增长分别为1.8％，1.2％和1.4％。
Training Deep Convolutional Neural Networks (CNNs) is based on the notion of using multiple kernels and non-linearities in their subsequent activations to extract useful features. The kernels are used as general feature extractors without specific correspondence to the target class. As a result, the extracted features do not correspond to specific classes. Subtle differences between similar classes are modeled in the same way as large differences between dissimilar classes. To overcome the class-agnostic use of kernels in CNNs, we introduce a novel method named Class Regularization that performs class-based regularization of layer activations. We demonstrate that this not only improves feature search during training, but also allows an explicit assignment of features per class during each stage of the feature extraction process. We show that using Class Regularization blocks in state-of-the-art CNN architectures for action recognition leads to systematic improvement gains of 1.8%, 1.2% and 1.4% on the Kinetics, UCF-101 and HMDB-51 datasets, respectively.
