Abstract | ||
---|---|---|
This paper introduces a new framework for image classification using local visual descriptors. The pipeline first performs a non-linear feature transformation on descriptors, then aggregates the results together to form image-level representations, and finally applies a classification model. For all the three steps we suggest novel solutions which make our approach appealing in theory, more scalable in computation, and transparent in classification. Our experiments demonstrate that the proposed classification method achieves state-of-the-art accuracy on the well-known PASCAL benchmarks. |
Year | DOI | Venue |
---|---|---|
2010 | 10.1007/978-3-642-15555-0_11 | ECCV (5) |
Keywords | Field | DocType |
super-vector coding,local image descriptors,proposed classification method,classification model,image-level representation,new framework,well-known pascal benchmarks,non-linear feature transformation,state-of-the-art accuracy,novel solution,local visual descriptors,image classification | Data mining,Computer science,Coding (social sciences),Vector quantization,Artificial intelligence,Visual descriptors,Contextual image classification,Computation,Feature transformation,Computer vision,Pattern recognition,Kullback–Leibler divergence,Machine learning,Scalability | Conference |
Volume | ISSN | ISBN |
6315 | 0302-9743 | 3-642-15554-5 |
Citations | PageRank | References |
255 | 11.86 | 18 |
Authors | ||
4 |
Name | Order | Citations | PageRank |
---|---|---|---|
Xi Zhou | 1 | 404 | 25.54 |
Yu, Kai | 2 | 4799 | 255.21 |
Zhang, Tong | 3 | 7126 | 611.43 |
Thomas S. Huang | 4 | 27815 | 2618.42 |