2021 Fiscal Year Annual Research Report
Deep forest for disaster monitoring with multi-temporal and multi-modal earth observation data
| Project/Area Number |
19K20309
|
|---|
| Research Institution | Institute of Physical and Chemical Research |
|---|
Principal Investigator |
Xia Junshi 国立研究開発法人理化学研究所, 革新知能統合研究センター, 研究員 (00830168)
|
|---|
| Project Period (FY) |
2019-04-01 – 2022-03-31
|
| Keywords | Damage mapping / Machine learning / Multi-source datasets |
|---|
| Outline of Annual Research Achievements |
The research achievements can be summarized as follows:
1. We have proposed novel deep neural network architectures, including image adaptation, feature adaptation, knowledge distillation, and self-training (SL) modules to solve the differ-modality learning problem.
2. We have proposed a new coarse-to-fine framework for detecting floods at a large scale. Multiple source satellite datasets, including the Gaofen (GF) series and Zhuhai-1 hyperspectral, were provided to detect and monitor the floods.
3. We have proposed self-paced positive-unlabeled learning (PU) to build damage assessment with very limited labeled data and many unlabeled data.
4. We have proposed an object detection method and a background-enhancement method for building damage mapping. The experiment using both background-enhanced samples and landslide-inducing information showed a great performance.
|
Research Products
(4 results)
