 This paper proposes a novel 3D convolutional neural network, CNN, approach for automatic classification of crops from spatio-temporal remote sensing images. The 3D-CNN framework was designed to preserve the full crop growth cycle while also incorporating temporal information. Additionally, an active learning strategy was introduced to further improve the accuracy of the model. Experiments showed that the proposed 3D-CNN outperformed existing methods demonstrating its potential for automated crop classification. This article was authored by Shunping Ji, Qi Zhang, An Jinsu, and others.