LLaVA-generated captions fused with image features via dual cross-attention improve remote sensing scene classification accuracy over the paper's own simple baselines.
Indexing of remote sensing images with different resolutions by multiple features,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2024 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Multimodal Remote Sensing Scene Classification Using VLMs and Dual-Cross Attention Networks
LLaVA-generated captions fused with image features via dual cross-attention improve remote sensing scene classification accuracy over the paper's own simple baselines.