CoDeRNet is proposed, a unified framework that preserves task-specific representations while enabling bidirectional exchange of complementary information and introduces a Confidence-Decomposed Routing (CoDeR) mechanism, which decomposes the routing condition into receiver-side need for support and sender-side reliability cues to guide selective transfer.
Abstract
Change Detection&Captioning (CDC) aims to jointly localize changed regions and describe their semantic content from bi-temporal remote sensing images. While recent approaches have demonstrated the effectiveness of unified modeling, they primarily rely on shared representations or jointly optimized architectures, leaving task-specific communication under heterogeneous supervision less explored. However, detection and captioning differ fundamentally in their supervision granularity, leading to heterogeneous representations with uneven reliability across tasks and spatial locations. In this paper, we reformulate CDC as a selective information routing problem between heterogeneous task branches, where the key challenge is to determine when and where information should be transferred. To this end, we propose CoDeRNet, a unified framework that preserves task-specific representations while enabling bidirectional exchange of complementary information. CoDeRNet introduces a Confidence-Decomposed Routing (CoDeR) mechanism, which decomposes the routing condition into receiver-side need for support and sender-side reliability cues to guide selective transfer. Experiments on LEVIR-MCI and WHU-CDC show that CoDeRNet achieves strong overall CDC performance, particularly improving caption generation while preserving competitive change localization. Extensive analyses further show that the benefits arise not merely from joint learning or generic cross-task communication, but from explicitly modeling selective cross-task information routing. These results highlight the importance of conditional information transfer for unified CDC under heterogeneous supervision.
This work introduces structure-aware entropy-based matching discrepancy, which jointly models feature uncertainty and structural coherence to ensure accurate feature adaptation between graphs and develops a domain-aware semi-hard negative sampling strategy that constructs informative contrastive sets by filtering unrel...
Ming-Can Wang, Jun-Chang Xin, Zhong-Ming Yao et al.· arXiv.org· 0 citations
A leave-one-domain-out (LODO) transfer learning protocol for edge classification on DyTAGs is formally established, and it is demonstrated that state-of-the-art self-supervised methods for dynamic graph learning perform poorly when transferred to unseen domains.
A compact explainability-driven framework for comparing encoders trained with classification-only and caption-generation supervision on CUB-200 provides a systematic basis for assessing learned representations and comparing alternative training and transfer-learning strategies.
Cross-graph transfer on text-attributed graphs (TAGs) is challenging because different graph domains often encode different structure-semantic relationships. Existing unified-encoder methods jointly train a single encoder using multiple source graphs, but such a unified encoder yields only marginal improvements over th...
Minuk Kim, Heasoo Hwang· Proceedings of the 35th ACM...· 0 citations
When labels are corrupted by noise, cross-modal hashing often fails to maintain reliable semantic neighborhoods and clear decision boundaries in the Hamming space. Existing robust hashing methods mainly improve the reliability of samples or pairwise supervision, while the class-level semantics contained in multi-label...
Fan Yang, Yuan-Zhi Zhao, Tong-Xuan Pei et al.· Proceedings of the 34th ACM...· 0 citations
Self-training with teacher-generated pseudo-labels has driven notable progress in cross-domain object detection. However, most existing methods primarily rely on output-level supervision, while overlooking the transfer of intermediate representations. Although feature distillation appears to be a natural solution, dire...
Rui-Qing Wang, Yang Xu, Zhi-Hui Wei et al.· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.