DCANet: Dense Context-Aware Network for Semantic Segmentation
This work addresses semantic segmentation for computer vision applications, presenting an incremental improvement in context modeling.
The authors tackled semantic segmentation by proposing a Dense Context-Aware (DCA) module to integrate local and global context information, achieving improved performance on datasets like PASCAL VOC 2012, Cityscapes, and ADE20K with concrete gains such as 82.1% mIoU on Cityscapes.
As the superiority of context information gradually manifests in advanced semantic segmentation, learning to capture the compact context relationship can help to understand the complex scenes. In contrast to some previous works utilizing the multi-scale context fusion, we propose a novel module, named Dense Context-Aware (DCA) module, to adaptively integrate local detail information with global dependencies. Driven by the contextual relationship, the DCA module can better achieve the aggregation of context information to generate more powerful features. Furthermore, we deliberately design two extended structures based on the DCA modules to further capture the long-range contextual dependency information. By combining the DCA modules in cascade or parallel, our networks use a progressive strategy to improve multi-scale feature representations for robust segmentation. We empirically demonstrate the promising performance of our approach (DCANet) with extensive experiments on three challenging datasets, including PASCAL VOC 2012, Cityscapes, and ADE20K.