Performance of Domain Adaptation Schemes in Video Action Recognition using Synthetic Data
Performance of Domain Adaptation Schemes in Video Action Recognition using Synthetic Data
复制标题
DOI：
10.1145/3531232.3531242
复制
发表时间：
2022-03
期刊：
影响因子：
--
通讯作者：
Hana Isoi;A. Takefusa;H. Nakada;M. Oguchi
中科院分区：
文献类型：
--
作者：
Hana Isoi;A. Takefusa;H. Nakada;M. Oguchi
从监控摄像头获得的视频有望在神经网络中进行处理，并用于室内监控和监视服务。然而，收集训练数据通常是机器学习中的一个问题，从隐私保护的角度来看，获得足够的训练数据是非常困难的，特别是对于上述服务。为了解决这个问题，可以采用合成数据而不是真实的数据来进行训练。然而，视频的域移位尚未被充分研究，因此尚未实现使用合成视频的真实的视频的高精度无监督学习。在这项研究中，我们创建了一个合成的视频数据集的动作分类，并表明，动作分类可以进行高精度，而不标记真实的视频数据，通过学习与此合成数据集。首先，我们创建一个OchaHouse数据集，用于室内监测和监视的行动分类。该数据集由记录房间中一个人的行为的真实的视频数据集OchaHouse-真实的和类似于它的合成视频数据集OchaHouse-Syn组成。第二，使用合成数据和具有域自适应方案的各种无监督学习方法来执行真实的视频的视频动作识别。结果表明，使用合成数据的学习能够在没有真实的数据标签的情况下对真实的数据进行高度准确的动作分类。数据集将被发布。
The videos obtained from surveillance cameras are expected to be processedin neuralnetworksandutilizedforindoor monitoringand surveillance services. However, collecting training data is generally an issue in machine learning, obtaining sufficient training data is very difficult from the viewpoint of privacy protection, especially for the abovementioned services. To address this issue, synthetic data instead of real data might be employed for training. However, the domain shift of video has not been sufficiently investigated, and therefore high-accuracy, unsupervised learning of real video using synthetic video has not been achieved. In this study, we create a synthetic video dataset for action classification and show that action classification can be performed with high accuracy without labeling real video data by learning with this synthetic dataset. First, we create an OchaHouse Dataset for action classification for indoor monitoring and surveillance. This dataset consists of a real video dataset OchaHouse-Real that records the behavior of one person in a room and a synthetic video dataset OchaHouse-Syn that resembles it. Second, video action recognition of real videos is performed using synthetic data and various unsupervised learning methods with domain adaptation schemes. The results show that learning with synthetic data enables highly accurate action classification of real data without a real data label. The dataset will be published.