SYSTEMS AND METHODS FOR AUTOMATED INFERENCE OF CHANGES IN SPACE-TIME IMAGES

摘要

Systems and Methods for Automated Inference of Spatio-Temporal Imaging Changes The present disclosure addresses the technical problem of allowing automated inference of changes in spatio-temporal imaging by leveraging the high-level robust features extracted from a convolution neural network (CNN). trained in varied contexts rather than data-dependent resource methods. Unsupervised grouping of high-level features eliminates the inconvenient requirement to label images. Since templates are not trained in any specific context, any image can be accepted. Real-time inference is allowed by a combination of unsupervised grouping and supervised classification. A cloud edge topology ensures real-time inference even when connectivity is unavailable by ensuring that updated classification models are organized on the edge. Creating a knowledge ontology based on adaptive learning enables the inference of an input image with varying levels of accuracy. Precision farming may be an application of the present disclosure.