Big Earth

Big Earth data leverages massive, multi-modal Earth observation datasets (including satellite imagery, soundscapes, and meteorological data) to address pressing environmental challenges through machine learning. Current research focuses on developing and applying advanced deep learning architectures, such as convolutional LSTMs, transformers, and YOLO models, for tasks like land cover classification, change detection, and soundscape mapping, often incorporating explainable AI techniques. This field is significantly impacting various sectors by enabling more accurate and timely monitoring of climate change impacts, agricultural practices, and biodiversity, ultimately supporting sustainable development goals.