Multimodal large language models (MLLMs) must balance local detail against scene context when interpreting ultra-high-resolution (UHR) remote sensing (RS) imagery within a limited visual-input budget. Existing selection-based methods either prune tokens and select patches through relevance scoring, or crop actively thr...
Yao Zhang, Peng-Yu Dai, Wei Guo et al.· 0 citations
GeoSeg-OV is proposed, which decouples auxiliary VFM features from visual-text matching and repurposes them as structural guidance for cost aggregation and decoding, and introduces Cost-Aware Decoding (CAD) to adaptively refine and fuse multi-scale semantic and structural guidance based on the current decoder context.
Rui-Zhong Liu, Ting-Zhang Luo, Zai-Yan Zhang et al.· 0 citations
Rapid post-disaster response requires timely, building-level information on whether structures remain intact, are damaged, or are destroyed. Post-event optical imagery, however, may be unavailable because of cloud, smoke, or darkness. The Bright Challenge evaluated all-weather building damage mapping from a submeter-re...
Hongruixuan Chen, He Huang, Haifeng Wang et al.· arXiv.org· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
