Shape Embedding and Knowledge Mining Network for Generalized Few-Shot Remote Sensing Segmentation

In recent years, generalized few-shot segmentation (GFSS) has received widespread attention from scholars by virtue of its superiority in low-data regimes. Most of the existing research focuses on natural image processing, and few studies have been devoted to the practical but challenging topic of remote sensing image (RSI) understanding. In this letter, we propose a shape embedding and knowledge mining network (SKNet) for generalized few-shot RSI segmentation. Specifically, the framework is divided into two key stages: 1) in the base class learning stage, shape representation embedding is introduced to enhance the network’s ability to perceive remote sensing objects. Simultaneously, we introduce the self-reconstruction constraint (SRC) to prevent new unseen classes from merging, thereby improving the representation uniqueness of these classes and 2) in the novel class learning stage, a base class knowledge mining (BCKM) mechanism is designed to update the prototypes of the novel class using the prototype representation of the base class, so as to enhance the discrimination ability of the network. We validated our methods on the adapted version of the OpenEarthMap and iSAID datasets. In comparison to existing GFSS methods, the proposed approach demonstrates an advancement.