Author(s)

Xiaomin Gao

Corresponding Author

Xiaomin Gao

Abstract

In recent years, in-depth learning has developed rapidly, which overturns the idea of algorithm design in the fields of speech recognition, image classification and text comprehension. Because of its strong feature extraction ability, in-depth learning is especially outstanding in the field of image recognition. However, there is not much combination between in-depth learning and video monitoring. Because the depth model has multi-layer network structure and the complexity of algorithm, it is time-consuming to train and update the model, it is difficult to meet the real-time requirement. This paper reviews the development history of in-depth learning, introduces the main models of in-depth learning at home and abroad in recent 10 years, discusses the target tracking algorithm based on in-depth learning, points out the advantages and disadvantages of each algorithm, and finally summarizes the existing problems and prospects in this field.

Keywords

In-depth learning, Video tracking, Convolution neural network, Recurrent neural network, Self-encoder