Weakly Supervised Spatial Deep Learning for Earth Image Segmentation Based on Imperfect Polyline Labels

Mathematics

• Image Segmentation 100%
• Training 78%
• Learning 66%
• Imagery 36%
• Water Management 21%
• Water Resources 18%
• Resource Management 18%
• Remote Sensing 18%
• Unsupervised Learning 18%
• Performance Model 18%
• Contiguity 17%
• Supervised Learning 17%
• Annotation 17%
• Rivers 16%
• Model 16%
• Computer Vision 14%
• Alignment 14%
• Framework 14%
• Baseline 13%
• Line segment 13%
• Update 11%
• Neural Networks 11%
• Refinement 10%
• Semantics 10%
• Evaluation 8%
• Class 3%

Engineering & Materials Science

• Image segmentation 69%
• Deep learning 65%
• Labels 62%
• Earth (planet) 61%
• Pixels 11%
• Earth sciences 10%
• Water resources 9%
• Semi-supervised learning 9%
• Volatile organic compounds 8%
• Labeled data 8%
• Unsupervised learning 8%
• Supervised learning 7%
• Remote sensing 7%
• Rivers 7%
• Deep neural networks 7%
• Computer vision 6%
• Availability 5%