Name
Title | ||
|---|---|---|
Fast Fine-grained Image Classification via Weakly Supervised Discriminative Localization. |
Abstract | ||
|---|---|---|
Fine-grained image classification is to recognize hundreds of subcategories in each basic-level category. Existing methods employ discriminative localization to find the key distinctions between similar subcategories. However, they generally have two limitations: 1) discriminative localization relies on region proposal methods to hypothesize the locations of discriminative regions, which are
<italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">time-consuming</italic>
and the
<italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">bottleneck</italic>
of improving classification speed and 2) the training of discriminative localization depends on object or part annotations which are heavily
<italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">labor-consuming</italic>
and the
<italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">obstacle</italic>
of marching toward practical application. It is highly challenging to address the two limitations
<italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">simultaneously</italic>
, while existing methods only focus on one of them. Therefore, we propose a weakly supervised discriminative localization approach (WSDL) for fast fine-grained image classification to address the two limitations at the same time, and its main advantages are: 1) multi-level attention guided localization learning is proposed to localize discriminative regions with different focuses automatically, without using object and part annotations, avoiding the labor consumption. Different level attentions focus on different characteristics of the image, which are complementary and boost classification accuracy and 2)
<inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$n$ </tex-math></inline-formula>
-pathway end-to-end discriminative localization network is proposed to improve classification speed, which simultaneously localizes multiple different discriminative regions for one image to boost classification accuracy, and shares full-image convolutional features generated by a region proposal network to accelerate the process of generating region proposals as well as reduce the computation of convolutional operation. Both are jointly employed to simultaneously improve classification speed and eliminate dependence on object and part annotations. Comparing with state-of-the-art methods on two widely used fine-grained image classification data sets, our WSDL approach achieves the best accuracy and the efficiency of classification. |
Year | DOI | Venue |
|---|---|---|
2017 | 10.1109/TCSVT.2018.2834480 | IEEE Trans. Circuits Syst. Video Techn. |
Keywords | Field | DocType |
Proposals,Object detection,Feature extraction,Training,Detectors,Computer vision,Machine learning | Object detection,Bottleneck,Obstacle,Data set,Pattern recognition,Computer science,Feature extraction,Artificial intelligence,Contextual image classification,Discriminative model,Machine learning,Computation | Journal |
Volume | Issue | ISSN |
abs/1710.01168 | 5 | 1051-8215 |
Citations | PageRank | References |
4 | 0.40 | 29 |
Authors | ||
3 | ||
Name | Order | Citations | PageRank |
|---|---|---|---|
| Xiangteng He | 1 | 58 | 5.75 |
| Yuxin Peng | 2 | 1122 | 74.90 |
| JunJie Zhao | 3 | 107 | 12.05 |