Recursive multi-model complementary deep fusion for robust salient object detection via parallel sub-networks. (January 2022)

• Record Type: Journal Article
• Authors: Wu, Zhenyu; Li, Shuai; Chen, Chenglizhao; Hao, Aimin; Qin, Hong
• Is Part Of: Pattern recognition. Volume 121(2022)
• Publication Date: 2022-01
• Subjects: Salient object detection -- Deep learning -- Multi-model fusion; Pattern perception -- Periodicals; Perception des structures -- Périodiques; Patroonherkenning; 006.4
• DOI: 10.1016/j.patcog.2021.108212
• Languages: English
• ISSNs: 0031-3203

Abstract

Highlights: We utilize parallel sub networks to automatically reveal saliency clues at different spatial levels. We propose an end-to-end salient object detection model that learns diversity saliency clues in an iterative manner. We also provide a novel selective fusion strategy to fuse multi-model saliency clues for a high-performance salient object detection. Abstract: Fully convolutional networks have shown outstanding performance in the salient object detection (SOD) field. The state-of-the-art (SOTA) methods have a tendency to become deeper and more complex, which easily homogenize their learned deep features, resulting in a clear performance bottleneck. In sharp contrast to the conventional "deeper" schemes, this paper proposes a "wider" network architecture which consists of parallel sub-networks with totally different network architectures. In this way, those deep features obtained via these two sub-networks will exhibit large diversity, which will have large potential to be able to complement with each other. However, a large diversity may easily lead to the feature conflictions, thus we use the dense short-connections to enable a recursively interaction between the parallel sub-networks, pursuing an optimal complementary status between multi-model deep features. Finally, all these complementary multi-model deep features will be selectively fused to make high-performance salient object detections. Extensive experiments on several famous benchmarks clearly demonstrate the superior performance, good generalization, and powerful learning ability of the proposed wider framework.