关键词: 深度置信网络, 文字检测, 人脸及表情识别, 遥感图像领域

Abstract: This paper firstly introduces the development of Deep Belief Network （DBN） based on theory foundation. Afterwards, the difference between deep network structure and shallow network structure is analyzed. Finally, the literature makes a study and analysis of DBN, in the field of text detection, facial and expression recognition, and remote sensing image classification by quoting multiple representative documents. Through a comprehensive introduction to the deep learning model DBN and deeply understanding the construction and practical application of DBN, it provides researchers with the idea of improving DBN and applying it to a wider emerging field in the future.

Key words: Deep Belief Network（DBN）, text detection, facial and expression recognition, remote sensing image field