Abstract | ||
---|---|---|
In this work we propose the method for a rather unexplored problem of computer vision - discriminatively trained dense surface normal estimation from a single image. Our method combines contextual and segment-based cues and builds a regressor in a boosting framework by transforming the problem into the regression of coefficients of a local coding. We apply our method to two challenging data sets containing images of man-made environments, the indoor NYU2 data set and the outdoor KITTI data set. Our surface normal predictor achieves results better than initially expected, significantly outperforming state-of-the-art. |
Year | DOI | Venue |
---|---|---|
2014 | 10.1007/978-3-319-10602-1_31 | COMPUTER VISION - ECCV 2014, PT V |
Keywords | Field | DocType |
Computer Vision,Ground Truth,Random Forest,Visual Word,Feature Representation | Computer vision,Data set,Regression,Computer science,Coding (social sciences),Ground truth,Boosting (machine learning),Artificial intelligence,Random forest,Normal,Visual Word | Conference |
Volume | ISSN | Citations |
8693 | 0302-9743 | 45 |
PageRank | References | Authors |
1.84 | 42 | 3 |
Name | Order | Citations | PageRank |
---|---|---|---|
Ladický L'ubor | 1 | 1015 | 44.54 |
Bernhard Zeisl | 2 | 75 | 2.57 |
Marc Pollefeys | 3 | 7671 | 475.90 |