Temporally Ordered Region-Token Mamba with Logit-Space Diffusion for Remote Sensing Change Detection

Researchers introduced BMD-CD, a remote sensing change-detection model combining temporally ordered region-token Mamba state-space modeling with a five-step logit-space Conditional Diffusion Decoder, according to an arXiv paper (2609.27149v1). BMD-CD reached F1 scores of 93.7%, 96.0%, 97.8%, and 99.0% on the LEVIR-CD, WHU-CD, DSIFN-CD, and CDD benchmarks, and improved 3-pixel Boundary-F1 to 87.7% on LEVIR-CD and 91.4% on WHU-CD. The full model requires 32.09 GFLOPs and 47 ms per 256 x 256 image pair, with zero-shot transfer demonstrated on ValaisCD and B-FLAIR-test.

arXiv:2609.27149v1 Announce Type: new Abstract: Remote sensing change detection requires both global reasoning across bitemporal images and precise localization of changed regions. However, dense attention is computationally expensive for high-resolution imagery, while conventional feature fusion and coarse decoding may inadequately separate genuine changes from appearance variations or preserve object boundaries. We present Bitemporal Mamba-Diffusion for Change Detection BMD-CD , which combines temporally structured state-space modeling with logit-space diffusion refinement. BMD-CD converts deep bitemporal features into region tokens and arranges them in explicit temporal partitions before bidirectional state-space propagation. Its Bitemporal Ordered Mamba Operator enables long-range cross-temporal interaction with linear sequence complexity, while Orthogonal Feature Disentanglement forms a change-oriented output and a complementary rotated output using learned pairwise rotations and unchanged-region consistency. Multiscale decoding then produces coarse change logits, which are refined through a five-step Conditional Diffusion Decoder operating directly in logit space. Experiments on LEVIR-CD, WHU-CD, DSIFN-CD, CDD, and S2Looking demonstrate strong performance across diverse change-detection settings. BMD-CD achieves F1 scores of 93.7%, 96.0%, 97.8%, and 99.0% on the four standard benchmarks and improves 3-pixel Boundary-F1 to 87.7% and 91.4% on LEVIR-CD and WHU-CD, respectively. The full model requires 32.09 GFLOPs and 47 ms per 256 x 256 image pair, while also showing zero-shot transfer to ValaisCD and B-FLAIR-test. Our code is available at https://github.com/Aparup2139/Public WACV/