SwinNet: Swin Transformer Drives Edge-Aware RGB-D and RGB-T Salient Object Detection

Convolutional neural networks (CNNs) are good at extracting contexture features within certain receptive fields, while transformers can model the global long-range dependency features. By absorbing advantage of transformer and merit CNN, Swin Transformer shows strong feature representation ability. Based on it, we propose a cross-modality fusion model, SwinNet , for RGB-D RGB-T salient object detection. It is driven by to extract hierarchical features, boosted attention mechanism bridge gap between two modalities, guided edge information sharp contour object. To be specific, two-stream encoder first extracts multi-modality then spatial alignment channel re-calibration module presented optimize intra-level clarify fuzzy boundary, edge-guided decoder achieves inter-level under guidance The proposed outperforms state-of-the-art models datasets, showing that it provides more insight into complementarity task.