A Generalizable and Accessible Approach to Machine Learning with Global Satellite Imagery

Abstract

Combining satellite imagery with machine learning (SIML) has the potential to address global challenges by remotely estimating socioeconomic and environmental conditions in data-poor regions, yet the resource requirements of SIML limit its accessibility and use. We show that a single encoding of satellite imagery can generalize across diverse prediction tasks (e.g. forest cover, house price, road length). Our method achieves accuracy competitive with deep neural networks at orders of magnitude lower computational cost, scales globally, delivers label super-resolution predictions, and facilitates characterizations of uncertainty. Since image encodings are shared across tasks, they can be centrally computed and distributed to unlimited researchers, who need only fit a linear regression to their own ground truth data in order to achieve state-of-the-art SIML performance.

References in corpus (13)

• Very Deep Convolutional Networks for Large-Scale Image Recognition
• Neural Architecture Search with Reinforcement Learning
• Remote Sensing Image Scene Classification: Benchmark and State of the Art
• Unsupervised Deep Feature Extraction for Remote Sensing Image Classification
• A Comprehensive Survey of Deep Learning in Remote Sensing: Theories, Tools and Challenges for the Community
• Dense semantic labeling of sub-decimeter resolution images with convolutional neural networks
• Do ImageNet Classifiers Generalize to ImageNet?
• Toward Deeper Understanding of Neural Networks: The Power of Initialization and a Dual View on Expressivity
• Flare Prediction Using Photospheric and Coronal Image Data
• Transfer Learning from Deep Features for Remote Sensing and Poverty Mapping
• Poverty Prediction with Public Landsat 7 Satellite Imagery and Machine Learning
• Least Squares Revisited: Scalable Approaches for Multi-class Prediction
• Convolutional Kitchen Sinks for Transcription Factor Binding Site Prediction