Spatial Locality-Aware Sparse Coding and Dictionary Learning

Jiang Wang, Junsong Yuan, Zhouyuan Chen, Ying Wu
Proceedings of the Asian Conference on Machine Learning, PMLR 25:491-505, 2012.

Abstract

Nonlinear encoding of SIFT features has recently shown good promise in image classification. This scheme is able to reduce the training complexity of the traditional bag-of-feature approaches while achieving better performance. As a result, it is suitable for large-scale image classification applications. However, existing nonlinear encoding methods do not explicitly consider the spatial relationship when encoding the local features, but merely leaving the spatial information used at a later stage, e.g. through the spatial pyramid matching, is largely inadequate. In this paper, we propose a joint sparse coding and dictionary learning scheme that take the spatial information into consideration in encoding. Our experiments on synthetic data and benchmark data demonstrate that the proposed scheme can learn a better dictionary and achieve higher classification accuracy.

Cite this Paper

BibTeX
@InProceedings{pmlr-v25-wang12a, title = {Spatial Locality-Aware Sparse Coding and Dictionary Learning}, author = {Wang, Jiang and Yuan, Junsong and Chen, Zhouyuan and Wu, Ying}, booktitle = {Proceedings of the Asian Conference on Machine Learning}, pages = {491--505}, year = {2012}, editor = {Hoi, Steven C. H. and Buntine, Wray}, volume = {25}, series = {Proceedings of Machine Learning Research}, address = {Singapore Management University, Singapore}, month = {04--06 Nov}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v25/wang12a/wang12a.pdf}, url = {https://proceedings.mlr.press/v25/wang12a.html}, abstract = {Nonlinear encoding of SIFT features has recently shown good promise in image classification. This scheme is able to reduce the training complexity of the traditional bag-of-feature approaches while achieving better performance. As a result, it is suitable for large-scale image classification applications. However, existing nonlinear encoding methods do not explicitly consider the spatial relationship when encoding the local features, but merely leaving the spatial information used at a later stage, e.g. through the spatial pyramid matching, is largely inadequate. In this paper, we propose a joint sparse coding and dictionary learning scheme that take the spatial information into consideration in encoding. Our experiments on synthetic data and benchmark data demonstrate that the proposed scheme can learn a better dictionary and achieve higher classification accuracy.} }
Endnote
%0 Conference Paper %T Spatial Locality-Aware Sparse Coding and Dictionary Learning %A Jiang Wang %A Junsong Yuan %A Zhouyuan Chen %A Ying Wu %B Proceedings of the Asian Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2012 %E Steven C. H. Hoi %E Wray Buntine %F pmlr-v25-wang12a %I PMLR %P 491--505 %U https://proceedings.mlr.press/v25/wang12a.html %V 25 %X Nonlinear encoding of SIFT features has recently shown good promise in image classification. This scheme is able to reduce the training complexity of the traditional bag-of-feature approaches while achieving better performance. As a result, it is suitable for large-scale image classification applications. However, existing nonlinear encoding methods do not explicitly consider the spatial relationship when encoding the local features, but merely leaving the spatial information used at a later stage, e.g. through the spatial pyramid matching, is largely inadequate. In this paper, we propose a joint sparse coding and dictionary learning scheme that take the spatial information into consideration in encoding. Our experiments on synthetic data and benchmark data demonstrate that the proposed scheme can learn a better dictionary and achieve higher classification accuracy.
RIS
TY - CPAPER TI - Spatial Locality-Aware Sparse Coding and Dictionary Learning AU - Jiang Wang AU - Junsong Yuan AU - Zhouyuan Chen AU - Ying Wu BT - Proceedings of the Asian Conference on Machine Learning DA - 2012/11/17 ED - Steven C. H. Hoi ED - Wray Buntine ID - pmlr-v25-wang12a PB - PMLR DP - Proceedings of Machine Learning Research VL - 25 SP - 491 EP - 505 L1 - http://proceedings.mlr.press/v25/wang12a/wang12a.pdf UR - https://proceedings.mlr.press/v25/wang12a.html AB - Nonlinear encoding of SIFT features has recently shown good promise in image classification. This scheme is able to reduce the training complexity of the traditional bag-of-feature approaches while achieving better performance. As a result, it is suitable for large-scale image classification applications. However, existing nonlinear encoding methods do not explicitly consider the spatial relationship when encoding the local features, but merely leaving the spatial information used at a later stage, e.g. through the spatial pyramid matching, is largely inadequate. In this paper, we propose a joint sparse coding and dictionary learning scheme that take the spatial information into consideration in encoding. Our experiments on synthetic data and benchmark data demonstrate that the proposed scheme can learn a better dictionary and achieve higher classification accuracy. ER -
APA
Wang, J., Yuan, J., Chen, Z. & Wu, Y.. (2012). Spatial Locality-Aware Sparse Coding and Dictionary Learning. Proceedings of the Asian Conference on Machine Learning, in Proceedings of Machine Learning Research 25:491-505 Available from https://proceedings.mlr.press/v25/wang12a.html.

Related Material