- Rating
- 1
- Confidence
- 2
- Informativeness
- 2
- Presentation
- 2
- Soundness
- 1
Read review
Paper summary
This paper addresses cross-resolution remote sensing change detection and adaptation to semantic classes held out during base-detector training. The proposed pipeline first reconstructs the post-event image using conditional diffusion super-resolution, then performs cross-temporal change detection. A frozen segmentation model generates candidate regions, which a PPO policy iteratively refines, accepts, or rejects using retrieval-augmented multimodal rewards. An external semantic model assigns pseudo-labels to accepted regions, and only the additional class head is trained. The authors claim improvements in cross-resolution change detection and semantic coverage without acquiring manual pixel-level annotations for the new classes in target scenes, while largely preserving performance on previously seen classes.
Critical strengths
- The application problems are meaningful, and several task boundaries are clearly specified. Resolution mismatch and semantic expansion after deployment are practically relevant challenges. The paper distinguishes resolution alignment from geometric registration and clarifies the role of external semantic knowledge, making the setting more precise than a broad claim of zero-shot discovery.
- The evaluation is broad, including component ablations, manual verification, and performance retention on seen classes.
Critical weaknesses
- The central contribution lacks focus. The pipeline combines several established components, but the connection between resolution alignment and adaptation to new classes, and the need for a unified framework, remain unclear.
- The comparison protocols are inconsistent. Table 2 states that baseline results come from different publications and may use different protocols, whereas the text claims identical inputs, splits, and training conditions.
- Recent baseline coverage is limited. Table 2 contains no 2026 methods. Please justify the selection and discuss task-comparable work available before manuscript completion.
- SR training conditions need clarification. Were SwinIR and OSEDiff in Table 2 fine-tuned on the same remote sensing data, and which checkpoints and budgets were used? Matching output dimensions and the detector does not control for domain-adaptation differences.
- The independent benefit of RL is insufficiently established. A greedy multi-step baseline with the same reward and model-call budget is missing. Acceptance precision/FDR alone cannot exclude improvements obtained by accepting fewer candidates.
Questions for authors
- Which baselines in Table 2 were rerun under matched conditions? Were the SR baselines in Table 4 fine-tuned on the same remote sensing data?
- What ground-truth labels and denominators were used for held-out-class coverage and post-adaptation F1/mIoU?
- Do you have results comparing RL with greedy refinement under the same reward and computational budget, including both precision and recall?