Abstract | ||
---|---|---|
A number of recent systems for unsupervised feature-based learning of object models take advantage of co-occurrence: broadly, they search for clusters of discriminative features that tend to coincide across multiple still images or video frames. An intuition behind these efforts is that regularly co-occurring image features are likely to refer to physical traits of the same object, while features that do not often co-occur are more likely to belong to different objects. In this paper we discuss a refinement to these techniques in which multiple segmentations establish meaningful contexts for co-occurrence, or limit the spatial regions in which two features are deemed to co-occur. This approach can reduce the variety of image data necessary for model learning and simplify the incorporation of less discriminative features into the model. |
Year | DOI | Venue |
---|---|---|
2005 | 10.1109/ACVMOT.2005.119 | WACV/MOTION |
Keywords | Field | DocType |
discriminative feature,multiple segmentation,image data,object model,physical trait,co-occurring image feature,recent system,unsupervised feature-based learning,different object,learn feature-based object models,meaningful context,feature extraction,learning artificial intelligence,image segmentation,image features | Scale-space segmentation,Feature detection (computer vision),Computer science,Segmentation-based object categorization,Image segmentation,Artificial intelligence,Discriminative model,Computer vision,Pattern recognition,Feature (computer vision),Feature extraction,Video tracking,Machine learning | Conference |
ISBN | Citations | PageRank |
0-7695-2271-8-1 | 1 | 0.40 |
References | Authors | |
12 | 2 |
Name | Order | Citations | PageRank |
---|---|---|---|
Thomas Stepleton | 1 | 10 | 1.66 |
Tai Sing Lee | 2 | 794 | 88.73 |