Abstract | ||
---|---|---|
Many methods have been proposed to solve the image classification problem for a large number of categories. Among them, methods based on tree-based representations achieve good trade-off between accuracy and test time efficiency. While focusing on learning a tree-shaped hierarchy and the corresponding set of classifiers, most of them [11, 2, 14] use a greedy prediction algorithm for test time efficiency. We argue that the dramatic decrease in accuracy at high efficiency is caused by the specific design choice of the learning and greedy prediction algorithms. In this work, we propose a classifier which achieves a better trade-off between efficiency and accuracy with a given tree-shaped hierarchy. First, we convert the classification problem as finding the best path in the hierarchy, and a novel branch-and-bound-like algorithm is introduced to efficiently search for the best path. Second, we jointly train the classifiers using a novel Structured SVM (SSVM) formulation with additional bound constraints. As a result, our method achieves a significant 4.65%, 5.43%, and 4.07% (relative 24.82%, 41.64%, and 109.79%) improvement in accuracy at high efficiency compared to state-of-the-art greedy "tree-based" methods [14] on Caltech-256 [15], SUN [32] and Image Net 1K [9] dataset, respectively. Finally, we show that our branch-and-bound-like algorithm naturally ranks the paths in the hierarchy (Fig. 8) so that users can further process them. |
Year | DOI | Venue |
---|---|---|
2013 | 10.1109/ICCV.2013.40 | ICCV |
Keywords | Field | DocType |
high efficiency,tree-shaped hierarchy,hierarchical classifier,best path,tree searching,large scale image classification,greedy prediction algorithm,tree-based representations,branch-and-bound,accurate classifier,image hierarchies,good trade-off,image classification,novel branch-and-bound-like algorithm,greedy algorithms,classification problem,ssvm formulation,image classification problem,test time efficiency,caltech-256,structured svm formulation,better trade-off,image net 1k,support vector machines,branch-and-bound-like algorithm,sun | Structured support vector machine,Branch and bound,Pattern recognition,Computer science,Support vector machine,Greedy algorithm,Artificial intelligence,Hierarchical classifier,Hierarchy,Greedy randomized adaptive search procedure,Contextual image classification | Conference |
Volume | Issue | ISSN |
2013 | 1 | 1550-5499 |
Citations | PageRank | References |
14 | 0.51 | 29 |
Authors | ||
3 |
Name | Order | Citations | PageRank |
---|---|---|---|
Min Sun | 1 | 1083 | 59.15 |
Wan Huang | 2 | 14 | 0.51 |
Silvio Savarese | 3 | 3975 | 161.69 |