基于改进3D ResNet的视频人体行为识别方法研究

A video human behavior recognition method based on improved 3D ResNet

针对人体行为在视频中呈现的时序性特点,提出了一种融合非对称卷积和CBR模块的视频人体行为识别方法。该方法使用3D ResNet-50作为主干网络。首先,将网络中较大的卷积核拆分为2个非对称3D卷积核的串联,加深卷积层在水平和竖直方向上的局部关键特征提取;其次,加入了CBR模块,以增加网络层数。该网络对连续的视频帧序列进行图像和时序的多角度特征提取,并根据特征数据对其进行分类,最后输出识别结果。在基准数据集UCF101上的大量实验结果表明,所提方法的Top1准确率和Top5准确率与原始3D ResNet网络相比分别提升了4.03%和4.99%,且该方法的识别准确率也优于其他主流方法的识别准确率。

Aiming at the temporal characteristics of human behavior in videos,a video human behavior recognition method is proposed that combines asymmetric convolution and CBR modules.This method uses 3D ResNet-50 as the backbone network.First,the larger convolutions in the network are changed to the concatenation of two asymmetric 3D convolutions,which deepens the local key feature extraction of the convolution layer in the horizontal and vertical directions.Secondly,CBR module is added to improve the number of network layers.The network extracts multi-angle features of images and time series from continuous video frame sequences,classifies them according to the feature data,and finally outputs the recognition results.Extensive experimental results on the benchmark dataset UCF101 show that the Top1 and Top5 accuracy of the proposed method are improved by 4.03%and 4.99%,respectively,compared with the original 3D ResNet network,and the recognition accuracy of this method is also better than other mainstream methods.