ProtoFlow: Mitigating Forgetting in Class-Incremental Remote Sensing Segmentation via Low-Curvature Prototype Flow
2026-04-03 • Computer Vision and Pattern Recognition
Computer Vision and Pattern Recognition
AI summaryⓘ
The authors address the problem of continuously updating remote sensing image segmentation as new categories appear and conditions change over time. They propose ProtoFlow, a method that tracks how typical examples (prototypes) of each class move and change over time, using a special model of their evolution. By keeping these prototype changes smooth and distinct between classes, their method reduces forgetting past knowledge and improves accuracy. Tests show better performance compared to other methods, suggesting this time-aware approach helps maintain stable recognition in changing environments.
remote sensingsemantic segmentationcontinual learningprototype learningrepresentation driftincremental learningtemporal vector fieldclass prototypesforgettingmIoU
Authors
Jiekai Wu, Rong Fu, Chuangqi Li, Zijian Zhang, Guangxin Wu, Hao Zhang, Shiyin Lin, Jianyuan Ni, Yang Li, Dongxu Zhang, Amir H. Gandomi, Simon Fong, Pengbin Feng
Abstract
Remote sensing segmentation in real deployment is inherently continual: new semantic categories emerge, and acquisition conditions shift across seasons, cities, and sensors. Despite recent progress, many incremental approaches still treat training steps as isolated updates, which leaves representation drift and forgetting insufficiently controlled. We present ProtoFlow, a time-aware prototype dynamics framework that models class prototypes as trajectories and learns their evolution with an explicit temporal vector field. By jointly enforcing low-curvature motion and inter-class separation, ProtoFlow stabilizes prototype geometry throughout incremental learning. Experiments on standard class- and domain-incremental remote sensing benchmarks show consistent gains over strong baselines, including up to 1.5-2.0 points improvement in mIoUall, together with reduced forgetting. These results suggest that explicitly modeling temporal prototype evolution is a practical and interpretable strategy for robust continual remote sensing segmentation.