A Domain-Specific Memory Bank-Enhanced Framework for Remote Sensing Image Captioning

Remote sensing image captioning enables automated interpretation of remote sensing data for applications such as environmental monitoring and urban planning. However, general-purpose multimodal models exhibit suboptimal adaptability on remote sensing imagery due to domain discrepancies in data distribution, resolution, and rotation invariance. We propose a domain-specific memory bank-enhanced framework for remote sensing image caption, built upon the lightweight EVCap architecture with three core innovations: (1) a remote sensing specific visual memory bank providing domain-representative visual-semantic alignment; (2) a staged transfer training strategy that leverages pre-trained knowledge while mitigating overfitting from limited remote sensing data; and (3) a K-means clustering-based few-shot memory bank that preserves performance while reducing memory footprint for resource-constrained deployment. Extensive experiments on NWPU, RSICD, and Sydney datasets demonstrate substantial improvements over baselines. Specifically, CIDEr increases by 3.3% on RSICD (0.2735 to 0.2825) and 11.7% on Sydney (0.1422 to 0.1589). The proposed method also achieves competitive performance against state-of-the-art methods while maintaining favorable computational efficiency.