X. Huang and G. Zhao, "Spontaneous facial micro-expression analysis using spatiotemporal local radon-based binary pattern," 2017 International Conference on the Frontiers and Advances in Data Science (FADS), Xi'an, 2017, pp. 159-164. doi: 10.1109/FADS.2017.8253219
Spontaneous facial micro-expression analysis using spatiotemporal local radon-based binary pattern
|Author:||Huang, Xiaohua1; Zhao, Guoying2,1|
1Center for Machine Vision and Signal Analysis, University of Oulu, Finland
2School of Information and Technology Northwest University, Xi’an, China
|Online Access:||PDF Full Text (PDF, 1.6 MB)|
|Persistent link:|| http://urn.fi/urn:nbn:fi-fe2019040511212
Institute of Electrical and Electronics Engineers,
|Publish Date:|| 2019-04-05
Micro-expressions are difficult to be observed by human beings due to its low intensity and short duration. Recently, several works have been developed to resolve the problems of micro-expression recognition caused by subtle intensity and short duration. One of them, Local binary pattern from three orthogonal planes (LBP-TOP) is primarily used to recognize micro-expression from the video recorded by high-speed camera. Several variances of LBP-TOP have also been developed to promisingly improve the performance of LBP-TOP for micro-expression recognition. However, these variances of LBP-TOP including LBP-TOP cannot well extract the subtle movements of micro-expression so that they have the low performance. In this paper, we propose spontaneous local radon-based binary pattern to analyze micro-expressions with subtle facial movements. Firstly, it extracts the sparse information by using robust principal component analysis since micro-expression data are sparse in both temporal and spatial domains caused by short duration and low intensity. These sparse information can provide much motion information to dynamic feature descriptor. Furthermore, it employs radon transform to obtain the shape features from three orthogonal planes, as radon transform is robustness to the same histogram distribution of two images. Finally, one-dimensional LBP is employed in these shape features for constructing the spatiotemporal features for micro-expression video. Intensive experiments are conducted on two available published micro-expression databases including SMIC and CASME2 databases for evaluating the performance of the proposed method. Experimental results demonstrate that the proposed method achieves promising performance in micro-expression recognition.
|Pages:||159 - 164|
2017 International Conference on the Frontiers and Advances in Data Science (FADS)
International Conference on the Frontiers and Advances in Data Science
|Type of Publication:||
A4 Article in conference proceedings
|Field of Science:||
113 Computer and information sciences
This work was supported by the Academy of Finland, Tekes Fidipro Program, the strategic Funds of the University of Oulu, Finland, the Infotech Oulu. We gratefully acknowledge the support of NVIDIA Corporation with the donation of the Titan X Pascal GPU used for this research.
© 2017 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works.