Abstract | ||
---|---|---|
This work proposes a general multi-layer framework for image labeling, which targets the challenging problem of classifying the occluded parts of the 3D scene depicted in a 2D image. Our framework is based on the mixed graphical models, which explicitly encode causal relationship between the visible and occluded regions. Unlike other image labeling techniques where a single label is determined for each pixel, layered model assigns multiple labels to pixels. We propose a novel “Multi-Layer-CRF” framework that allows for the integration of sophisticated occlusion potentials into the model and enables the automatic inference of the layer decomposition. We use a special message-passing algorithm to perform maximum a posterior inference on mixed graphs and demonstrate the ability to infer the correct labels of occluded regions in both the aerial near-vertical dataset and urban street-view dataset. It is shown to increase the classification accuracy in occluded areas significantly. |
Year | DOI | Venue |
---|---|---|
2019 | 10.1007/s11042-018-6298-5 | Multimedia Tools Appl. |
Keywords | Field | DocType |
Conditional random fields, Graphical models, Classification, Semantic segmentation, Occlusions | Conditional random field,ENCODE,Layered model,Computer vision,Multi layer,Pattern recognition,Inference,Computer science,Mixed graph,Artificial intelligence,Pixel,Graphical model | Journal |
Volume | Issue | ISSN |
78 | 2 | 1573-7721 |
Citations | PageRank | References |
1 | 0.34 | 22 |
Authors | ||
3 |
Name | Order | Citations | PageRank |
---|---|---|---|
Sergey Kosov | 1 | 4 | 1.11 |
Kimiaki Shirahama | 2 | 108 | 22.43 |
Marcin Grzegorzek | 3 | 185 | 48.00 |