The rapid development of multimodal large language models (MLLMs) has highlighted the need for comprehensive benchmarks to assess their capabilities and limitations in remote sensing (RS). However, evaluating MLLMs in large-size synthetic aperture radar (SAR) scenarios remains challenging, as understanding complete scenes requires both global semantic interpretation and fine-grained target perception. Existing SAR benchmarks predominantly rely on small image patches, limiting their ability to assess these capabilities jointly. To address these limitations, we present XLSAR-Bench, a comprehensive benchmark for evaluating the perception and reasoning capabilities of MLLMs in large-scene SAR imagery. XLSAR-Bench comprises 405 images with an average size of 7153$\times$7153 pixels, ten tasks spanning visual question answering, visual grounding, and detailed image captioning, with all evaluation samples meticulously annotated manually, assisted by a semi-automatic captioner. Experimental results from general-purpose and domain-specific MLLMs reveal persistent challenges in understanding large SAR scenes and accurately perceiving small targets. We have open-sourced XLSAR-Bench to support further research in developing more powerful MLLMs for complete large SAR scenes.
翻译:暂无翻译