RMMamba: Randomized Mamba for Remote Sensing Shadow Removal

Remote sensing shadow removal task aims to effec­tively restore key regions of an image obscured by shadows. However, the spatial nonuniformity of shadow distribution presents significant challenges to this task. To address this issue, we propose RMMamba, a shadow removal network based on the SS2D architecture. The core concept of RMMamba involves balancing the spatial nonuniformity of shadow distribution, thereby optimizing the utilization efficiency of nonshadow pixel information across different windows. Specifically, RMMamba employs a random pixel shuffling operation to evenly disperse pixels from shadow regions with pronounced spatial nonunifor­mity into nonshadow areas, ensuring a more balanced spatial distribution of shadow and nonshadow pixels within each window. Subsequently, a shared-weight local state-space model (SSM) (SS2D) is employed to integrate nonshadow pixel features uni­formly distributed around shadow pixels, consequently effectively relighting shadow pixels. Reverse shuffling operations are then applied to restore the processed image to its original pixel order. Coupled with color prior feedforward network (CP-FFN), a lightweight feedforward network incorporating color priors, RMMamba effectively restores color in shadow regions. More importantly, given the difficulty of acquiring remote sensing shadow samples with corresponding ground truth and shadow masks, we leverage the game Grand Theft Auto (GTA) to control its shadow renderer and create shadow removal grand theft auto (SRGTA), a synthetic fully supervised dataset, hence providing a new benchmark for the performance evaluation of remote sensing shadow removal algorithms. Extensive experiments conducted on SRGTA and UAV image dataset for shadow correction (UAV-SC) have demonstrated the outstanding performance of RMMamba.