SARChat-Bench-2M: A Multi-Task Vision-Language Benchmark for SAR Image Interpretation
This addresses the lack of domain-specific datasets for SAR image interpretation, benefiting military, maritime, and infrastructure monitoring applications, though it is incremental as it adapts existing VLM frameworks to a new domain.
The paper introduces SARChat-2M, a large-scale multimodal dialogue dataset with 2 million image-text pairs for SAR images, enabling and evaluating vision-language models in SAR image interpretation tasks such as visual understanding and object detection.
As a powerful all-weather Earth observation tool, synthetic aperture radar (SAR) remote sensing enables critical military reconnaissance, maritime surveillance, and infrastructure monitoring. Although Vision language models (VLMs) have made remarkable progress in natural language processing and image understanding, their applications remain limited in professional domains due to insufficient domain expertise. This paper innovatively proposes the first large-scale multimodal dialogue dataset for SAR images, named SARChat-2M, which contains approximately 2 million high-quality image-text pairs, encompasses diverse scenarios with detailed target annotations. This dataset not only supports several key tasks such as visual understanding and object detection tasks, but also has unique innovative aspects: this study develop a visual-language dataset and benchmark for the SAR domain, enabling and evaluating VLMs' capabilities in SAR image interpretation, which provides a paradigmatic framework for constructing multimodal datasets across various remote sensing vertical domains. Through experiments on 16 mainstream VLMs, the effectiveness of the dataset has been fully verified. The project will be released at https://github.com/JimmyMa99/SARChat.