Abstract | ||
---|---|---|
We propose a cluster-based approach to point set saliency detection, a challenge since point sets lack topological information. A point set is first decomposed into small clusters, using fuzzy clustering. We evaluate cluster uniqueness and spatial distribution of each cluster and combine these values into a cluster saliency function. Finally, the probabilities of points belonging to each cluster are used to assign a saliency to each point. Our approach detects fine-scale salient features and uninteresting regions consistently have lower saliency values. We evaluate the proposed saliency model by testing our saliency-based keypoint detection against a 3D interest point detection benchmark. The evaluation shows that our method achieves a good balance between false positive and false negative error rates, without using any topological information. |
Year | DOI | Venue |
---|---|---|
2015 | 10.1109/ICCV.2015.27 | ICCV |
Keywords | Field | DocType |
point set saliency detection,cluster-based approach,fuzzy clustering,cluster uniqueness evaluation,spatial distribution evaluation,probabilities,fine-scale salient feature detection,uninteresting region detection,saliency-based key-point detection,3D interest point detection,false positive error rates,false negative error rates | Cluster (physics),Computer vision,Fuzzy clustering,Uniqueness,Kadir–Brady saliency detector,Pattern recognition,Computer science,Salience (neuroscience),Interest point detection,Artificial intelligence,Point set,Salient | Conference |
Volume | Issue | ISSN |
2015 | 1 | 1550-5499 |
Citations | PageRank | References |
9 | 0.45 | 24 |
Authors | ||
3 |
Name | Order | Citations | PageRank |
---|---|---|---|
Flora Ponjou Tasse | 1 | 53 | 4.41 |
Jirí Kosinka | 2 | 84 | 17.76 |
Neil A. Dodgson | 3 | 723 | 54.20 |