Review on the Architectures and Applications of Deep Learning in Agriculture

This article is a brief review of the architectures of the state-of-the-art deep learning networks, including Deep Belief Networks (DBNs), AutoEncoders (AEs), Convolutional Neural Networks (CNNs) and Recurrent Neural Networks (RNNs) and their applications in agriculture. The paper is organized as follows. First, a brief introduction to some basic concepts is conducted to clarify the widely spoken terms. Then, main deep learning architectures are introduced, and characteristics of them are compared briefly. Thirdly, by surveying related literatures in the applications of the above architectures in agriculture, the typical and proper applications of each structure are summarized and analyzed. Last but not least, the challenges and outlook of deep learning networks are discussed.