Problem
Research questions and friction points this paper is trying to address.
class-incremental learning
catastrophic forgetting
remote sensing segmentation
continual learning
representation drift
ProtoFlow: Mitigating Forgetting in Class-Incremental Remote Sensing Segmentation via Low-Curvature Prototype Flow
🤖 AI Summary
This work addresses catastrophic forgetting and representation drift in remote sensing image segmentation caused by continuously expanding semantic classes and environmental changes. It introduces, for the first time, a time-aware prototype dynamics framework that models class prototypes as temporal trajectories and explicitly learns their evolution within a low-curvature vector field. By jointly optimizing the smoothness of prototype motion and inter-class separability, the proposed low-curvature prototype flow mechanism effectively stabilizes the geometric structure of feature representations during incremental learning. Evaluated on standard remote sensing incremental segmentation benchmarks, the method achieves a 1.5–2.0 percentage point improvement in mIoU<sub>all</sub>, significantly reduces forgetting, and successfully balances semantic stability with adaptability.
📝 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.
Innovation
Methods, ideas, or system contributions that make the work stand out.
prototype dynamics
low-curvature flow
class-incremental learning
remote sensing segmentation
continual learning
🔎 Similar Papers
