摘要
projection can be utilized to divide 360 degrees images into multiple rectilinear images, with little distortion. However, the existing saliency prediction models fail to integrate semantic information of these images. In this paper, we address this by proposing an intra-and inter-reasoning graph convolutional network for saliency prediction on 360 degrees images (SalReGCN360). The whole framework contains six subnetworks, each of which contains two branches. In the training phase, after utilizing Multiple Cubic Projection (MCP), six rectilinear images are simultaneously put into corresponding sub-networks. In one of the branches, the global features of a single rectilinear image are extracted by the intra-graph inference module to finely predict local saliency of 360 degrees images. In the other branch, the contextual features are extracted by the inter-graph inference module to effectively integrate semantic information of six rectilinear images. Finally, the feature maps are generated by the two branches fusion, and six corresponding rectilinear saliency maps are predicted. Extensive experiments on two popular saliency datasets illustrate the superiority of the proposed model, especially the improvement in KLD metric.
学科与关键词
- SalBiNet360: Saliency Prediction on 360° Images with Local-Global Bifurcated Deep NetworkChen, Dongwen; Qing, Chunmei*; Xu, Xiangmin; Zhu, Huansheng27th IEEE Conference on Virtual Reality and 3D User Interfaces (IEEE VR), Atlanta, GA, 2020-03-22 To 2020-03-26.被引 7
- MULTI-SCALE TRANSFORMER NETWORK FOR SALIENCY PREDICTION ON 360-DEGREE IMAGESLin, Xu; Qing, Chunmei*; Tan, Junpeng; Xu, Xiangmin30th IEEE International Conference on Image Processing (ICIP), Kuala Lumpur, MALAYSIA, 2023-10-08 To 2023-10-11.
- Attentive and context-aware deep network for saliency prediction on omni-directional imagesQing, Chunmei; Zhu, Huansheng; Xing, Xiaofen*; Chen, Dongwen; Jin, JianxiuDigital Signal Processing: A Review Journal , 2022, 120: 103289.被引 4
