Asst. Prof. Luka Čehovin Zajc, University of Ljubljana, Faculty of Computer and Information Science
Автор: Nacionalni kompetenčni center SLING
Загружено: 2025-12-03
Просмотров: 12
Deep learning for Earth observation and Geoinformatics - Asst. Prof. Luka Čehovin Zajc, University of Ljubljana, Faculty of Computer and Information Science; Visual Cognitive Systems Lab
The explosive growth of Earth observation data presents major challenges for analysis at scale. Deep learning has emerged as a powerful approach, but practitioners face persistent bottlenecks, from limited labeled data to the high computational cost of training and inference. We present research directions to address these issues, including pretraining on massive unlabeled datasets, scalable training strategies, and our work towards efficient deployment of models in geospatial applications.
Доступные форматы для скачивания:
Скачать видео mp4
-
Информация по загрузке:
