Abstract | ||
---|---|---|
This paper presents a novel approach for labeling objects based on multiple spatially-registered images of a scene. We argue
that such a multi-view labeling approach is a better fit for applications such as robotics and surveillance than traditional
object recognition where only a single image of each scene is available. To encourage further study in the area, we have collected
a data set of well-registered imagery for many indoor scenes and have made this data publicly available. Our multi-view labeling
approach is capable of improving the results of a wide variety of image-based classifiers, and we demonstrate this by producing
scene labelings based on the output of both the Deformable Parts Model of [1] as well as a method for recognizing object contours
which is similar to chamfer matching. Our experimental results show that labeling objects based on multiple viewpoints leads
to a significant improvement in performance when compared with single image labeling.
|
Year | DOI | Venue |
---|---|---|
2010 | 10.1007/978-3-642-19315-6_36 | Asian Conference on Computer Vision |
Keywords | Field | DocType |
multiple viewpoint recognition,traditional object recognition,object contour,novel approach,deformable parts model,multiple viewpoint,better fit,multiple spatially-registered image,indoor scene,single image,object recognition | Computer vision,Null Object pattern,Pattern recognition,Chamfer matching,Viewpoints,Computer science,Artificial intelligence,Robotics,Cognitive neuroscience of visual object recognition,Image labeling | Conference |
Volume | ISSN | Citations |
6492 | 0302-9743 | 12 |
PageRank | References | Authors |
0.85 | 22 | 5 |
Name | Order | Citations | PageRank |
---|---|---|---|
Scott Helmer | 1 | 176 | 11.49 |
David Meger | 2 | 403 | 32.90 |
Marius Muja | 3 | 133 | 5.85 |
James J. Little | 4 | 2430 | 269.59 |
D. G. Lowe | 5 | 15718 | 1413.60 |