Search papers, labs, and topics across Lattice.
The Adaptive Triangular Transformer for Cloud Removal (ATT-CR) addresses the challenges of high computational complexity and interference from cloudy pixels in existing Transformer-based methods for cloud removal in remote sensing images. By introducing Triangular Attention (TAN) to approximate Softmax attention with O(N) complexity and a Feature Selected Gating Module (FSGM) to differentiate between cloudy and clean features, ATT-CR significantly enhances both efficiency and accuracy. Experimental results show that ATT-CR outperforms current state-of-the-art methods on cloud removal benchmarks, making it a promising solution for this task.
ATT-CR cuts computational costs while boosting cloud removal accuracy by effectively distinguishing between valid and invalid pixel information.
Cloud removal aims to accurately reconstruct the ground objects obscured by clouds in remote sensing images. Existing Transformer-based methods utilizing self-attention have shown impressive results by effectively modeling long-range dependencies in cloudy images. However, they suffer from the following issues: 1) the high computational complexity of self-attention limits scalability; 2) treating both cloudy and clean pixels as valid within the attention computation brings disturbances in subsequent layers, leading to suboptimal performance. To address these challenges, we propose the Adaptive Triangular Transformer for Cloud Removal (ATT-CR), a model that effectively reduces computational costs and mitigates interference from cloudy pixels. Specifically, it consists of two core components: Triangular Attention (TAN) and Feature Selected Gating Module (FSGM). TAN employs lower and upper triangular matrices to approximate Softmax attention with O(N) computational complexity, significantly reducing the computational costs. The FSGM, on the other hand, integrates with TAN to adaptively distinguish between cloudy and clean features, which minimizes the introduction of invalid information into subsequent layers. Extensive experiments on cloud removal benchmarks demonstrate that ATT-CR delivers superior performance compared to existing methods.
