ESIFCD: A Dual-Temporal Cross-Domain Fusion Network for Remote Sensing Change Detection

Remote sensing change detection (RSCD) must remain reliable under illumination, seasonal, and viewpoint variability while handling objects that range from narrow roads to block-scale complexes. We present Efficient Spectral Inception Former Change Detector (<italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">ESIFCD</i>), an efficient dual-temporal network that couples domain-aligned differencing with multiple cross-resolution exchange and lightweight global context encoding. ESIFCD introduces: (i) an <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Adaptive Difference Fusion Unit</i> (ADFU) that fuses absolute and learned differences over spatial, spectral, and texture features; (ii) a <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Cross-Resolution Aggregation Module</i> (CRAM) that repeatedly exchanges evidence among four scales to mitigate scale disagreement; (iii) a <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Patch-Aware Context Encoder</i> (PACE) that models long-range dependencies via tokenized patches with modest cost; and (iv) a fast <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">GLCM-Stats Encoder</i> that supplies texture semantics (ASM/Contrast/Correlation) without analytical histograms. A re-parameterized large-kernel block (RLKB) sharpens local detail, and a late <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">MetaCNNGate</i> learns pixel-wise weights over one primary and four auxiliary heads to consolidate their complementary expertise (small vs. large structures). Across six public benchmarks, ESIFCD attains the best F1 on each dataset: 93.19% (LEVIR-CD), 87.89% (LEVIR-CD+), 70.10% (S2Looking), 95.95% (CDD), 83.48% (SYSU-CD), and 95.30% (WHU-CD). The model is efficient and fast (12.33 GFLOPs; 82.07 images/s at 2562), and with automatic mixed precision reaches over 165 images/s while maintaining accuracy, surpassing recent strong baselines under the same setting. Ablations confirm that ADFU improves robustness to radiometric and registration noise, CRAM preserves thin structures and prevents fragmentation on large objects, PACE improves cluttered scenes, and MetaCNNGate consistently lifts F1 on all benchmarks.