Semantic segmentation via sparse coding over hierarchical regions

摘要: The purpose of this paper is segmenting objects in an image and assigning a predefined semantic label to each object. There are two contributions paper. On one hand, segmentation guided by hierarchical regions instead single-level or multi-scale generated multiple segmentations. the other sparse coding introduced as high level description regions, which contributes reduction quantization error compared traditional bag-of-visual-words method. Experiments on challenging Microsoft Research Cambridge dataset (MSRC 21) show that our algorithm achieves state-of-the-art performance.