Abstract
A selection of pooling region is an important issue on image classification process. Although Spatial Pyramid Matching (SPM) simply selects pooling regions by partitioning an image into rectangular grids, the SPM has been used the most until now to exploit the spatial information of features. In this paper, in addition to the SPM, we concatenate a novel pooling method which decomposes an image into an object region and a background region. Since the regions deal with not only the spatial information but also objectness that is higher level knowledge than the simple rectangular grids, flexible sub-regions can be achieved with our method. In order to estimate object region from an unlabeled image, we adopt the visual saliency map for a probabilistic measure of objectness. To do this, our soft region pooling replaces the existing max pooling, which helps features inside the probabilistic boundary of object/background region to be selected. Beyond applying the objectness score only to select features for the max pooling, weight for each element of image representation vector is also applied to emphasize an important component vector. The proposed method is implemented based on LLC, and tested on widely used dataset, Caltech-101 and Caltech-256. The proposed method achieves 2.77% improvement on Caltech-101 and achieves 1.84% improvement on Caltech-256 compared with the accuracy of LLC.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
References
Csurka, G., Dance, C., Fan, L., Willamowski, J., Bray, C.: Visual categorization with bags of keypoints. In: Workshop on ECCV 1, pp. 1–2 (2004)
Sivic, J., Zisserman, A.: Video Google: A text retrieval approach to object matching in videos. In: Proceedings of ICCV, vol. 2, pp. 1470–1477 (2003)
Lowe, D.G.: Object recognition from local scale-invariant features. In: ICCV, pp. 1150–1157 (1999)
Dalal, N., Triggs, B.: Histograms of oriented gradients for human detection. In: CVPR (1), pp. 886–893 (2005)
Wang, J., Yang, J., Yu, K., Lv, F., Huang, T.S., Gong, Y.: Locality-constrained linear coding for image classification. In: CVPR, pp. 3360–3367 (2010)
Perronnin, F., Sánchez, J., Mensink, T.: Improving the fisher kernel for large-scale image classification. In: Daniilidis, K., Maragos, P., Paragios, N. (eds.) ECCV 2010, Part IV. LNCS, vol. 6314, pp. 143–156. Springer, Heidelberg (2010)
Yang, J., Yu, K., Gong, Y., Huang, T.S.: Linear spatial pyramid matching using sparse coding for image classification. In: CVPR, pp. 1794–1801 (2009)
Coates, A., Ng, A.Y., Lee, H.: An analysis of single-layer networks in unsupervised feature learning. In: AISTATS, pp. 215–223 (2011)
Jia, Y., Huang, C., Darrell, T.: Beyond spatial pyramids: Receptive field learning for pooled image features. In: CVPR, pp. 3370–3377 (2012)
Lazebnik, S., Schmid, C., Ponce, J.: Beyond bags of features: Spatial pyramid matching for recognizing natural scene categories. In: CVPR, pp. 2169–2178 (2006)
Russakovsky, O., Lin, Y., Yu, K., Fei-Fei, L.: Object-centric spatial pooling for image classification. In: Fitzgibbon, A., Lazebnik, S., Perona, P., Sato, Y., Schmid, C. (eds.) ECCV 2012, Part II. LNCS, vol. 7573, pp. 1–15. Springer, Heidelberg (2012)
Nilsback, M.E., Zisserman, A.: A visual vocabulary for flower classification. In: CVPR (2), pp. 1447–1454 (2006)
Fei-Fei, L., Fergus, R., Perona, P.: Learning generative visual models from few training examples: An incremental bayesian approach tested on 101 object categories (2004)
Griffin, G., Holub, A., Perona, P.: Caltech-256 object category dataset. Technical Report 7694, California Institute of Technology (2007)
Alexe, B., Deselaers, T., Ferrari, V.: What is an object? In: CVPR, pp. 73–80 (2010)
Jiang, P., Ling, H., Yu, J., Peng, J.: Salient region detection by ufo: Uniqueness, focusness and objectness. In: ICCV, pp. 1976–1983 (2013)
Jiang, B., Zhang, L., Lu, H., Yang, C., Yang, M.H.: Saliency detection via absorbing markov chain. In: ICCV, pp. 1665–1672 (2013)
Fisher, R.A.: The use of multiple measurements in taxonomic problems. Annals of Eugenics 7, 179–188 (1936)
Feng, J., Ni, B., Tian, Q., Yan, S.: Geometric;p-norm feature pooling for image classification. In: CVPR, pp. 2697–2704 (2011)
Bosch, A., Zisserman, A., Munoz, X.: Image classification using random forests and ferns. In: ICCV (2007)
Fan, R.E., Chang, K.W., Hsieh, C.J., Wang, X.R., Lin, C.J.: LIBLINEAR: A library for large linear classification. JMLR 9, 1871–1874 (2008)
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2014 Springer International Publishing Switzerland
About this paper
Cite this paper
Heo, B., Jeong, H., Kim, J., Choi, SI., Choi, J.Y. (2014). Weighted Pooling Based on Visual Saliency for Image Classification. In: Bebis, G., et al. Advances in Visual Computing. ISVC 2014. Lecture Notes in Computer Science, vol 8887. Springer, Cham. https://doi.org/10.1007/978-3-319-14249-4_62
Download citation
DOI: https://doi.org/10.1007/978-3-319-14249-4_62
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-14248-7
Online ISBN: 978-3-319-14249-4
eBook Packages: Computer ScienceComputer Science (R0)