A5035527545- Remote Sensing in Agriculture
- Land Use and Ecosystem Services
- Multimodal Machine Learning Applications
- Synthetic Aperture Radar (SAR) Applications and Techniques
- Remote-Sensing Image Classification
- COVID-19 diagnosis using AI
- Species Distribution and Climate Change
- Domain Adaptation and Few-Shot Learning
- Flood Risk Assessment and Management
- Soil erosion and sediment transport
China National Commission for Disaster Reduction
2023
Peking University
2019-2023
Timely and accurate wetland information is necessary for resource management. Recent advances in machine learning remote sensing have facilitated cost-effective monitoring of wetlands. However, reliable methods fine-grained rapid mapping are still lacking. To address the issue, a sample set with 20 categories China was collected based on sampling strategy that combines automatic generation visual interpretation. Simultaneously, novel multi-stage method classification proposed, which...
Semi-supervised learning (SSL) essentially pursues class boundary exploration with less dependence on human annotations. Although typical attempts focus ameliorating the inevitable error-prone pseudo-labeling, we think differently and resort to exhausting informative semantics from multiple probably correct candidate labels. In this paper, introduce Fuzzy Positive Learning (FPL) for accurate SSL semantic segmentation in a plug-and-play fashion, targeting adaptively encouraging fuzzy positive...
Irregular spatial dependency is one of the major characteristics remote sensing images, which brings about challenges for classification tasks. Deep supervised models such as convolutional neural networks (CNNs) have shown great capacity image classification. However, they generally require a huge labeled training set fine tuning deep network. To handle irregular images and mitigate conflict between limited samples demand, we design superpixel-guided layer-wise embedding CNN (SLE-CNN)...
Accurate information on forest distribution is an essential basis for the protection of resources. Recent advances in remote sensing and machine learning have contributed to monitoring forest-cover cost-effectively, but reliable methods rapid mapping over mountainous areas are still lacking. In addition, landscape pattern has proven be closely related functioning ecosystems, yet few studies explicitly measured or revealed its driving forces areas. To address these challenges, we developed a...
Timely cropland information is crucial for ensuring food security and promoting sustainable development. Traditional field survey methods are time-consuming costly, making it difficult to support rapid monitoring of large-scale changes. Furthermore, most existing studies focus on evaluation from a single aspect such as quantity or quality, thus cannot comprehensively reveal spatiotemporal characteristics cropland. In this study, method evaluating the quality using multi-source remote...