Abstract
Monitoring changes triggered by mining activities is crucial for industrialcontrolling, environmental management and regulatory compliance, yet it posessignificant challenges due to the vast and often remote locations of miningsites. Remote sensing technologies have increasingly become indispensable todetect and analyze these changes over time. We thus introduce MineNetCD, acomprehensive benchmark designed for global mining change detection usingremote sensing imagery. The benchmark comprises three key contributions. First,we establish a global mining change detection dataset featuring more than 70kpaired patches of bi-temporal high-resolution remote sensing images andpixel-level annotations from 100 mining sites worldwide. Second, we develop anovel baseline model based on a change-aware Fast Fourier Transform (ChangeFFT)module, which enhances various backbones by leveraging essential spectrumcomponents within features in the frequency domain and capturing thechannel-wise correlation of bi-temporal feature differences to learnchange-aware representations. Third, we construct a unified change detection(UCD) framework that integrates over 13 advanced change detection models. Thisframework is designed for streamlined and efficient processing, utilizing thecloud platform hosted by HuggingFace. Extensive experiments have been conductedto demonstrate the superiority of the proposed baseline model compared with 12state-of-the-art change detection approaches. Empirical studies on modularizedbackbones comprehensively confirm the efficacy of different representationlearners on change detection. This contribution represents significantadvancements in the field of remote sensing and change detection, providing arobust resource for future research and applications in global miningmonitoring. Dataset and Codes are available via the link.