Now showing items 1-2 of 2
Improving semantic segmentation with generalized models of local context
(Springer International Publishing AG, 2017)
Semantic segmentation (i.e. image parsing) aims to annotate each image pixel with its corresponding semantic class label. Spatially consistent labeling of the image requires an accurate description and modeling of the local ...
Kernel kod-tablosu kodlaması ile sahne etiketleme
Superpixel based methods have recently shown success in scene segmentation and labeling. In scene labeling, a superpixel algorithm is used first to segment the image into visually consistent small regions; then several ...