Building Cluster-Class Association for Detecting paddy fields under Semi-Supervised deep learning framework

Semi-supervised learning is a common training paradigm used in the field of remote sensing, utilising the advantage of unsupervised learning to improve upon supervised models that have less labeled data for training. An orchestration of two learning methods is proposed by exploiting cluster-class associations. The architecture has two parts, in the front-end, a CNN is trained to perform clustering. The response of clustering becomes input to the classifier, back-end of the architecture. The classifier is a selectively connected neural network, where every node in the hidden layer represents a class. The whole architecture is then trained with a limited set of labeled data. During training, the weights of the front-end architecture are not updated. We apply our algorithm to detect paddy fields using Sentinel-l SAR data from 2018 and 2019.