Progressive multi-scale fusion network for RGB-D salient object detection

Salient object detection (SOD) aims at locating the most significant within a given image. In recent years, great progress has been made in applying SOD on many vision tasks. The depth map could provide additional spatial prior and boundary cues to boost performance. Combining information with image data obtained from standard visual cameras widely used works, however, introducing suboptimal fusion strategy may have negative influence performance of SOD. this paper, we discuss about advantages so-called progressive multi-scale method propose mask-guided feature aggregation module (MGFA). proposed framework can effectively combine two features different modalities and, furthermore, alleviate impact erroneous features, which are inevitably caused by variation quality. We further introduce refinement (MGRM) complement high-level semantic reduce irrelevant fusion, leading an overall detection. Experiments five challenging benchmarks demonstrate that outperforms 11 state-of-the-art methods under evaluation metrics. • Novel structure for RGB-D saliency Mask-Guided Feature Aggregation filtering noise data. Refinement Module RGB Progressive deep shallow layers. Achieve competitive compared prevalent methods.