Search papers, labs, and topics across Lattice.
This paper introduces a two-stage framework for point-supervised change detection (PS-CD) that leverages SAM2 priors to enhance the quality of pseudo-labels generated from sparse point annotations. In Stage I, the method employs a bi-temporal mask selection strategy to refine generic segmentation responses into reliable change pseudo-labels, followed by a lightweight CNN refinement module that incorporates an uncertainty-aware loss for improved boundary quality. Stage II implements a teacher-student self-training approach, where the teacher's pseudo-labels are periodically updated, leading to significant performance improvements over existing weakly supervised methods across multiple benchmark datasets.
Point-supervised change detection can achieve performance on par with fully supervised methods by effectively refining noisy pseudo-labels through a novel two-stage optimization process.
Point-supervised change detection (PS-CD) aims to identify pixel-level changes between bi-temporal images using only sparsely annotated points. Although point annotations substantially reduce labeling costs, their limited spatial coverage often results in incomplete and noisy pseudo-labels. To address this issue, we propose a two-stage framework that introduces SAM2 priors into PS-CD and progressively adapts them to the target task. In Stage I, SAM2 generates object-aware candidate masks from point annotations on the bi-temporal images, and a bi-temporal mask selection strategy is designed to convert generic segmentation responses into more reliable change pseudo-labels. Subsequently, a lightweight CNN refinement module with an uncertainty-aware loss is employed to improve boundary quality and local structural consistency. In Stage II, we construct a teacher-student self-training framework in which the teacher is updated by exponential moving average and periodically refreshes the pseudo-labels. This design establishes a closed-loop optimization process that alternates between pseudo-label refinement and model re-optimization. Experiments on three benchmark datasets, including WHU-CD, LEVIR-CD, and SYSU-CD, demonstrate that the proposed method outperforms previous weakly supervised approaches on most benchmarks and remains competitive with several fully supervised methods.
