Accurate and timely prediction of crop yield based on remote sensing data is important for food security. However, crop growth is a complex process, which makes it quite difficult to achieve better performance. To address this problem, a novel 3D convolutional neural multi-kernel network (3DMKGP) is proposed to capture hierarchical features for predicting crop yield. First, a full 3D convolutional neural network (3D CNN) is constructed to maximally explore deep spatial-spectral features from multispectral images. Then, a multi-kernel learning (MKL) approach is proposed for fusion of intra-image deep spatialspectral features and inter-sample spatial consistency features. Specifically, we assign a group of nonlinear kernels for each feature in the MKL framework, which provides a robust way to fit features extracted from different domains. Finally, the probability distribution of prediction results is obtained by a kernel based-method. We evaluate the performance of the proposed method on China wheat yield prediction and offer detailed and systematic analyses of the performance of the proposed method. In addition, our method is compared with several competing methods. Experimental results demonstrate that the proposed method has certain advantages and can provide better prediction performance than the competitive methods.
In recent years, convolutional neural networks (CNNs) have made great achievements in object extraction from very high-resolution (VHR) images. However, most existing approaches require large quantities of clean and accurate training data to achieve impressive classification results. The presence of inaccurate labels in training datasets is known to deteriorate the performance of CNNs. In this paper, we introduce a novel efficient method for improving the robustness when training CNN on the dataset with relatively noisy labels. First, we propose a feature and label noise model (FLNM) to model the noisy label distribution in the training dataset. Then, we use a multitask deep learning framework (MDLF) to integrate the FLNM into the training process of CNN. Finally, a novel loss function concerning the high-level features is introduced to efficiently train the MDLF. We evaluate our method on datasets from Massachusetts and compare this method with other state-of-the-art methods. The experimental results demonstrate the effectiveness of the proposed method in improving the classification performance of CNNs trained with noisy training dataset.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.