Categories | Descriptions | Methods |
---|---|---|
Supervised Learning | The corresponding labeled disaster training samples are available in the study area | MFFENet |
Semi-supervised Learning | Few labeled disaster training samples are available in the study area | SSCDNet, FixMatch |
Unsupervised Change Detection | There are no corresponding labeled samples in the research area, but there are pre-disaster remote sensing images | UCDFormer, DCVA |
Unsupervised Domain Adaptation with Source Data | Domain adaptation with source data module problem by leveraging the adversarial learning behaviors of GANs to perform distribution alignment in the pixel, feature, and output spaces | ADANet, CaGAN |
Unsupervised Domain Adaptation without Source Data | Domain adaptation without source data problem by generating a reliable synthetic source domain | SGD-MA |
Xu Q, Shi Y, Zhu X X. DisasterNets: Embedding Machine Learning in Disaster Mapping[J]. arXiv preprint arXiv:2306.09815, 2023.