Class incremental learning of remote sensing images based on class similarity distillation

• Published in: PeerJ. Computer science 2023-09, Vol.9, p.e1583-e1583, Article e1583
• Main Authors: Shen, Mingge, Chen, Dehu, Hu, Silan, Xu, Gang
• Format: Article
• Language: English
• Subjects: Analysis; Artificial Intelligence; Catastrophic forgetting; Class incremental learning; Class similarity distillation; Computer Vision; Global similarity distillation; Neural Networks; Remote sensing
• ISSN: 2376-5992; 2376-5992
• DOI: 10.7717/peerj-cs.1583

When a well-trained model learns a new class, the data distribution differences between the new and old classes inevitably cause catastrophic forgetting in order to perform better in the new class. This behavior differs from human learning. In this article, we propose a class incremental object detection method for remote sensing images to address the problem of catastrophic forgetting caused by distribution differences among different classes. First, we introduce a class similarity distillation (CSD) loss based on the similarity between new and old class prototypes, ensuring the model’s plasticity to learn new classes and stability to detect old classes. Second, to better extract class similarity features, we propose a global similarity distillation (GSD) loss that maximizes the mutual information between the new class feature and old class features. Additionally, we present a region proposal network (RPN)-based method that assigns positive and negative labels to prevent mislearning issues. Experiments demonstrate that our method is more accurate for class incremental learning on public DOTA and DIOR datasets and significantly improves training efficiency compared to state-of-the-art class incremental object detection methods.