Correlation-based burstiness for logo retrieval
Authors:
Open Access
- 29 October 2012
- conference paper
- Published by Association for Computing Machinery (ACM)
- p. 965-968
Abstract
International audienceDetecting logos in photos is challenging. A reason is that logos locally resemble patterns frequently seen in random images. We propose to learn a statistical model for the distribution of incorrect detections output by an image matching algorithm. It results in a novel scoring criterion in which the weight of correlated keypoint matches is reduced, penalizing irrelevant logo detections. In experiments on two very diff erent logo retrieval benchmarks, our approach largely improves over the standard matching criterion as well as other state-of-the-art approachesKeywords
This publication has 13 references indexed in Scilit:
- Scalable logo recognition in real-world images2011
- Interactive visual object search through mutual information maximization2010
- Spatial pyramid mining for logo detection in natural scenes2008
- Hamming Embedding and Weak Geometric Consistency for Large Scale Image Search2008
- Trademark matching and retrieval in sports video databases2007
- Total Recall: Automatic Query Expansion with a Generative Feature Model for Object Retrieval2007
- Modeling word burstiness using the Dirichlet distribution2005
- Distinctive Image Features from Scale-Invariant KeypointsInternational Journal of Computer Vision, 2004
- Distribution of content words and phrases in text and language modellingNatural Language Engineering, 1996
- SLINK: An optimally efficient algorithm for the single-link cluster methodThe Computer Journal, 1973