AutoBD: Automated Bi-Level Description for Scalable Fine-Grained Visual Categorization
- 13 September 2017
- journal article
- research article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Image Processing
- Vol. 27 (1) , 10-23
Abstract
Compared with traditional image classification, fine-grained visual categorization is a more challenging task, because it targets to classify objects belonging to the same species, e.g., classify hundreds of birds or cars. In the past several years, researchers have made many achievements on this topic. However, most of them are heavily dependent on the artificial annotations, e.g., bounding boxes, part annotations, and so on. The requirement of artificial annotations largely hinders the scalability and application. Motivated to release such dependence, this paper proposes a robust and discriminative visual description named Automated Bi-level Description (AutoBD). “Bi-level” denotes two complementary part-level and object-level visual descriptions, respectively. AutoBD is “automated,” because it only requires the image-level labels of training images and does not need any annotations for testing images. Compared with the part annotations labeled by the human, the image-level labels can be easily acquired, which thus makes AutoBD suitable for large-scale visual categorization. Specifically, the part-level description is extracted by identifying the local region saliently representing the visual distinctiveness. The object-level description is extracted from object bounding boxes generated with a co-localization algorithm. Although only using the image-level labels, AutoBD outperforms the recent studies on two public benchmark, i.e., classification accuracy achieves 81.6% on CUB-200-2011 and 88.9% on Car-196, respectively. On the large-scale Birdsnap data set, AutoBD achieves the accuracy of 68%, which is currently the best performance to the best of our knowledge.Keywords
Funding Information
- National Nature Science Foundation of China (61525206, 61572050, 91538111, 61429201, 61428207)
- Beijing Advanced Innovation Center for Imaging Technology (BAICIT-2016009)
- ARO (W911NF-15-1-0290)
- Faculty Research Gift Awards by NEC Laboratories of America and Blippar
This publication has 29 references indexed in Scilit:
- Caffe2014
- Revisiting the Fisher vector for fine-grained classificationPattern Recognition Letters, 2014
- 3D Object Representations for Fine-Grained Categorization2013
- POOF: Part-Based One-vs.-One Features for Fine-Grained Categorization, Face Verification, and Attribute Estimation2013
- Selective Search for Object RecognitionInternational Journal of Computer Vision, 2013
- Decision Forests: A Unified Framework for Classification, Regression, Density Estimation, Manifold Learning and Semi-Supervised LearningFoundations and Trends in Computer Graphics and Vision, 2012
- TriCoS: A Tri-level Class-Discriminative Co-segmentation Method for Image Classification2012
- Improving the Fisher Kernel for Large-Scale Image Classification2010
- The Pascal Visual Object Classes (VOC) ChallengeInternational Journal of Computer Vision, 2009
- Histograms of Oriented Gradients for Human Detection2005