Towards Spatial Variability Aware Deep Neural Networks (SVANN): A\n Summary of Results

2020/11/17 by Jayant Gupta, Yiqun Xie, Gupta, Jayant +3
Environmental Science · #Computer Vision and Pattern Recognition (cs.CV) #FOS: Computer and information sciences #Land Use and Ecosystem Services #Machine Learning (cs.LG) #Remote Sensing in Agriculture #Urban Green Space and Health

paper · pdf · doi:10.48550/arxiv.2011.08992

openalex publication_date 2020/11/17 · openalex created_date 2022/07/25 · openalex updated_date 2026/07/28

Abstract

Spatial variability has been observed in many geo-phenomena including\nclimatic zones, USDA plant hardiness zones, and terrestrial habitat types\n(e.g., forest, grasslands, wetlands, and deserts). However, current deep\nlearning methods follow a spatial-one-size-fits-all(OSFA) approach to train\nsingle deep neural network models that do not account for spatial variability.\nIn this work, we propose and investigate a spatial-variability aware deep\nneural network(SVANN) approach, where distinct deep neural network models are\nbuilt for each geographic area. We evaluate this approach using aerial imagery\nfrom two geographic areas for the task of mapping urban gardens. The\nexperimental results show that SVANN provides better performance than OSFA in\nterms of precision, recall,and F1-score to identify urban gardens.\n

Citations

• Artificial intelligence: a modern approach
• ImageNet classification with deep convolutional neural networks

Related

• Towards Comparative Physical Interpretation of Spatial Variability Aware Neural Networks: A Summary of Results
• A Subsampling Based Neural Network for Spatial Data
• DeepKriging: Spatially Dependent Deep Neural Networks for Spatial Prediction
• An Interpretable Implicit-Based Approach for Modeling Local Spatial Effects: A Case Study of Global Gross Primary Productivity
• Spatial regression graph convolutional neural networks: A deep learning paradigm for spatial multivariate distributions