MFEAFN: Multi-scale feature enhanced adaptive fusion network for image semantic segmentation.
Low-level features contain spatial detail information, and high-level features contain rich semantic information. Semantic segmentation research focuses on fully acquiring and effectively fusing spatial detail with semantic information. This paper proposes a multiscale feature-enhanced adaptive fusi...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Public Library of Science (PLoS)
2022-01-01
|
| Series: | PLoS ONE |
| Online Access: | https://doi.org/10.1371/journal.pone.0274249 |
| _version_ | 1798021170088378368 |
|---|---|
| author | Shusheng Li Liang Wan Lu Tang Zhining Zhang |
| author_facet | Shusheng Li Liang Wan Lu Tang Zhining Zhang |
| author_sort | Shusheng Li |
| collection | DOAJ |
| description | Low-level features contain spatial detail information, and high-level features contain rich semantic information. Semantic segmentation research focuses on fully acquiring and effectively fusing spatial detail with semantic information. This paper proposes a multiscale feature-enhanced adaptive fusion network named MFEAFN to improve semantic segmentation performance. First, we designed a Double Spatial Pyramid Module named DSPM to extract more high-level semantic information. Second, we designed a Focusing Selective Fusion Module named FSFM to fuse different scales and levels of feature maps. Specifically, the feature maps are enhanced to adaptively fuse these features by generating attention weights through a spatial attention mechanism and a two-dimensional discrete cosine transform, respectively. To validate the effectiveness of FSFM, we designed different fusion modules for comparison and ablation experiments. MFEAFN achieved 82.64% and 78.46% mIoU on the PASCAL VOC2012 and Cityscapes datasets. In addition, our method has better segmentation results than state-of-the-art methods. |
| first_indexed | 2024-04-11T17:09:27Z |
| format | Article |
| id | doaj.art-b45856729c994eeba77129412ff5ccee |
| institution | Directory Open Access Journal |
| issn | 1932-6203 |
| language | English |
| last_indexed | 2024-04-11T17:09:27Z |
| publishDate | 2022-01-01 |
| publisher | Public Library of Science (PLoS) |
| record_format | Article |
| series | PLoS ONE |
| spelling | doaj.art-b45856729c994eeba77129412ff5ccee2022-12-22T04:12:57ZengPublic Library of Science (PLoS)PLoS ONE1932-62032022-01-01179e027424910.1371/journal.pone.0274249MFEAFN: Multi-scale feature enhanced adaptive fusion network for image semantic segmentation.Shusheng LiLiang WanLu TangZhining ZhangLow-level features contain spatial detail information, and high-level features contain rich semantic information. Semantic segmentation research focuses on fully acquiring and effectively fusing spatial detail with semantic information. This paper proposes a multiscale feature-enhanced adaptive fusion network named MFEAFN to improve semantic segmentation performance. First, we designed a Double Spatial Pyramid Module named DSPM to extract more high-level semantic information. Second, we designed a Focusing Selective Fusion Module named FSFM to fuse different scales and levels of feature maps. Specifically, the feature maps are enhanced to adaptively fuse these features by generating attention weights through a spatial attention mechanism and a two-dimensional discrete cosine transform, respectively. To validate the effectiveness of FSFM, we designed different fusion modules for comparison and ablation experiments. MFEAFN achieved 82.64% and 78.46% mIoU on the PASCAL VOC2012 and Cityscapes datasets. In addition, our method has better segmentation results than state-of-the-art methods.https://doi.org/10.1371/journal.pone.0274249 |
| spellingShingle | Shusheng Li Liang Wan Lu Tang Zhining Zhang MFEAFN: Multi-scale feature enhanced adaptive fusion network for image semantic segmentation. PLoS ONE |
| title | MFEAFN: Multi-scale feature enhanced adaptive fusion network for image semantic segmentation. |
| title_full | MFEAFN: Multi-scale feature enhanced adaptive fusion network for image semantic segmentation. |
| title_fullStr | MFEAFN: Multi-scale feature enhanced adaptive fusion network for image semantic segmentation. |
| title_full_unstemmed | MFEAFN: Multi-scale feature enhanced adaptive fusion network for image semantic segmentation. |
| title_short | MFEAFN: Multi-scale feature enhanced adaptive fusion network for image semantic segmentation. |
| title_sort | mfeafn multi scale feature enhanced adaptive fusion network for image semantic segmentation |
| url | https://doi.org/10.1371/journal.pone.0274249 |
| work_keys_str_mv | AT shushengli mfeafnmultiscalefeatureenhancedadaptivefusionnetworkforimagesemanticsegmentation AT liangwan mfeafnmultiscalefeatureenhancedadaptivefusionnetworkforimagesemanticsegmentation AT lutang mfeafnmultiscalefeatureenhancedadaptivefusionnetworkforimagesemanticsegmentation AT zhiningzhang mfeafnmultiscalefeatureenhancedadaptivefusionnetworkforimagesemanticsegmentation |