1/1
2 files

Cascaded one-vs-rest detection network for ne-grained recognition without part annotation

journal contribution
posted on 23.04.2018, 10:18 by Long Chen, Shengke Wang, Kin-Man Lam, Huiyu Zhou, Muwei Jian, Junya Dong
Fine-grained recognition is a challenging task due to small intra-category variances. Most of the top-performing fine-grained recognition methods leverage parts of objects for better performance. Therefore, part annotations which are extremely computationally expensive are required. In this paper, we propose a novel cascaded deep CNN detection framework for fine-grained recognition which is trained to detect a whole object without considering parts. Nevertheless, most of the current top-performing detection networks use N + 1 class (N object categories plus background) softmax loss. The background category with much more training samples dominates the feature learning progress where the features are not suitable for object categorisation with fewer samples. To address this issue, we here introduce two strategies: 1) We leverage a cascaded structure to eliminate the background. 2) We introduce a novel one-vs-rest loss function to capture more minute variances from different subordinate categories. Experiments show that our proposed recognition framework achieves comparable performance against the state-of-the-art, part-free, fine-grained recognition methods on the CUB-200-2011 Bird dataset. Meanwhile, our method outperforms most of the existing part annotation based methods and does not need part annotations at the training stage whilst being free from any annotations at the test stage.

History

Citation

Multimedia Tools and Applications, 2018

Author affiliation

/Organisation/COLLEGE OF SCIENCE AND ENGINEERING/Department of Informatics

Version

AM (Accepted Manuscript)

Published in

Multimedia Tools and Applications

Publisher

Springer Verlag

issn

1380-7501

eissn

1573-7721

Acceptance date

09/02/2018

Copyright date

2018

Available date

17/03/2019

Publisher version

https://link.springer.com/article/10.1007/s11042-018-5875-y

Notes

The file associated with this record is under embargo until 12 months after publication, in accordance with the publisher's self-archiving policy. The full text may be available through the publisher links provided above.

Language

en

Usage metrics

Categories

Keywords

Exports