摘要
针对以往动态场景分类中需要手动提取动态特征描述符以及特征维数过高的问题,提出利用深度学习网络模型进行动态纹理特征的提取。首先利用慢特征分析法(SFA)预先学习每个视频序列的动态特征,将该特征作为深度学习网络模型的输入数据进行学习,进一步得到信号的高级表示,深度网络模型选用堆栈降噪自动编码模型,最后用SVM分类法对其进行分类。实验证明该方法所提取的特征维数低,并且能够有效地表示动态纹理。
To overcome the shortcomings of extracting the feature descriptors by manual operation and too high feature di?mension for dynamic scene classification,a deep learning network model is proposed to extract dynamic texture features. First?ly,the slow feature analysis method is used to learn dynamic characteristics of each video sequence through before hand,and the learned feature is used as input data of deep learning to get the advanced representation of the input signal. The stacked de?noising autoencoding model is selected for the deep learning network mode. SVM classification method is used for its classifica?tion. The experimental result proves that the feature dimension extracted by this method is low and can effectively describe dy?namic textures.
出处
《现代电子技术》
北大核心
2015年第6期20-24,共5页
Modern Electronics Technique
基金
国家自然科学基金(11204109)
江苏省高校自然科学基金(12KJB510003)
关键词
动态纹理分类
慢特征分析
深度学习
堆栈降噪自动编码网络模型
dynamic texture classification
slow feature analysis
deep learning
stacked denoising autoencoding model