Computer Electronic and Software Society (CESS) organized a workshop on Zero-Shot Learning for Remote Sensing Applications workshop at the Jinniuhu Campus on May 18th and 19th, 2024.
On the first day, the facilitator provided an introduction to Remote Sensing (RS) scenes, explaining how they are generated and the types of RS sensors used. Participants were enlightened on various types of RS downstream tasks, including scene classification, semantic segmentation, change detection, and more. The facilitator also introduced Contrastive Language Image Pre-training (CLIP).
On the second day, participants implemented two Zero-Shot Learning algorithms using two different types of Baseline models. Additionally, plans were made for future workshops in the same domain, with a focus on Feature Extraction models in a broader context of RS applications. The day concluded with a group picture capturing the enthusiasm of all participants.
