Technical Report on Subspace Pyramid Fusion Network for Semantic Segmentation
This work addresses semantic segmentation, a key task in computer vision, but appears incremental as it builds on existing multi-scale feature fusion approaches.
The paper tackles the problem of capturing multi-scale feature representations for semantic segmentation by proposing a Subspace Pyramid Fusion Module (SPFM) and an Efficient Shuffle Attention Module (ESAM) to reconstruct skip-connections. Experimental results on Camvid and Cityscapes datasets demonstrate the effectiveness of the method.
The following is a technical report to test the validity of the proposed Subspace Pyramid Fusion Module (SPFM) to capture multi-scale feature representations, which is more useful for semantic segmentation. In this investigation, we have proposed the Efficient Shuffle Attention Module(ESAM) to reconstruct the skip-connections paths by fusing multi-level global context features. Experimental results on two well-known semantic segmentation datasets, including Camvid and Cityscapes, show the effectiveness of our proposed method.