Earth Observation
Earth observation leverages satellite and aerial imagery to monitor and analyze Earth's surface, aiming to understand environmental changes and support sustainable development. Current research heavily utilizes deep learning, employing transformer and convolutional neural network architectures (like U-Nets and variations) for tasks such as land cover classification, disaster monitoring, and crop yield prediction, often incorporating multimodal data fusion (e.g., combining optical and radar imagery). These advancements improve the accuracy and efficiency of Earth observation data analysis, impacting various fields including agriculture, climate change research, and resource management.
Papers
May 23, 2022
April 14, 2022