Research Article
Improved U-Net-Like Network for Visual Saliency Detection Based on Pyramid Feature Attention
Table 1
Description of the six used benchmark datasets.
| | Dataset | Year | Publication | Image amount | Object area (%) | Annotation |
| | SOD | 2010 | CVPR-W | 300 | | Pixel-wise object level | | ECSSD | 2015 | CVPR-W | 1,000 | | Pixel-wise object level | | DUT-OMRON | 2013 | CVPR | 5,168 | | Pixel-wise object level | | PASCAL-S | 2014 | CVPR | 850 | | Pixel-wise object level | | HKU-IS | 2015 | CVPR | 4,447 | | Pixel-wise object level | | DUTS | 2017 | CVPR | 15,572 | | Pixel-wise object level |
|
|
