Limits...
Robust facial expression recognition via compressive sensing.

Zhang S, Zhao X, Lei B - Sensors (Basel) (2012)

Bottom Line: The CS theory is used to construct a sparse representation classifier (SRC).The effectiveness and robustness of the SRC method is investigated on clean and occluded facial expression images.Three typical facial features, i.e., the raw pixels, Gabor wavelets representation and local binary patterns (LBP), are extracted to evaluate the performance of the SRC method.

View Article: PubMed Central - PubMed

Affiliation: School of Physics and Electronic Engineering, Taizhou University, Taizhou 318000, China. tzczsq@163.com

ABSTRACT
Recently, compressive sensing (CS) has attracted increasing attention in the areas of signal processing, computer vision and pattern recognition. In this paper, a new method based on the CS theory is presented for robust facial expression recognition. The CS theory is used to construct a sparse representation classifier (SRC). The effectiveness and robustness of the SRC method is investigated on clean and occluded facial expression images. Three typical facial features, i.e., the raw pixels, Gabor wavelets representation and local binary patterns (LBP), are extracted to evaluate the performance of the SRC method. Compared with the nearest neighbor (NN), linear support vector machines (SVM) and the nearest subspace (NS), experimental results on the popular Cohn-Kanade facial expression database demonstrate that the SRC method obtains better performance and stronger robustness to corruption and occlusion on robust facial expression recognition tasks.

No MeSH data available.


Related in: MedlinePlus

Recognition results of different methods with reduced dimension of Gabor wavelets representation.
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC3376615&req=5

f4-sensors-12-03747: Recognition results of different methods with reduced dimension of Gabor wavelets representation.

Mentions: When using the raw pixels (i.e., the resized images of 32 × 32 pixels) and LBP features for experiments, the corresponding recognition results and standard deviations (std) of different methods, including NN, SVM, NS, as well as SRC, are given in Table 1. The recognition results of different methods along with reduced dimension of Gabor wavelets representation are presented in Figure 4. Table 2 shows the best accuracy of different methods with the corresponding reduced dimension of Gabor wavelets representation. The results in Tables 1–2 and Figure 4 reveal that SRC achieves an accuracy of 94.76% with the raw pixels, 97.14% with LBP features, and 98.1% at best with 50 reduced dimension of Gabor wavelets representation, outperforming the other used methods. This confirms the validity and high performance of SRC for facial expression recognition.


Robust facial expression recognition via compressive sensing.

Zhang S, Zhao X, Lei B - Sensors (Basel) (2012)

Recognition results of different methods with reduced dimension of Gabor wavelets representation.
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC3376615&req=5

f4-sensors-12-03747: Recognition results of different methods with reduced dimension of Gabor wavelets representation.
Mentions: When using the raw pixels (i.e., the resized images of 32 × 32 pixels) and LBP features for experiments, the corresponding recognition results and standard deviations (std) of different methods, including NN, SVM, NS, as well as SRC, are given in Table 1. The recognition results of different methods along with reduced dimension of Gabor wavelets representation are presented in Figure 4. Table 2 shows the best accuracy of different methods with the corresponding reduced dimension of Gabor wavelets representation. The results in Tables 1–2 and Figure 4 reveal that SRC achieves an accuracy of 94.76% with the raw pixels, 97.14% with LBP features, and 98.1% at best with 50 reduced dimension of Gabor wavelets representation, outperforming the other used methods. This confirms the validity and high performance of SRC for facial expression recognition.

Bottom Line: The CS theory is used to construct a sparse representation classifier (SRC).The effectiveness and robustness of the SRC method is investigated on clean and occluded facial expression images.Three typical facial features, i.e., the raw pixels, Gabor wavelets representation and local binary patterns (LBP), are extracted to evaluate the performance of the SRC method.

View Article: PubMed Central - PubMed

Affiliation: School of Physics and Electronic Engineering, Taizhou University, Taizhou 318000, China. tzczsq@163.com

ABSTRACT
Recently, compressive sensing (CS) has attracted increasing attention in the areas of signal processing, computer vision and pattern recognition. In this paper, a new method based on the CS theory is presented for robust facial expression recognition. The CS theory is used to construct a sparse representation classifier (SRC). The effectiveness and robustness of the SRC method is investigated on clean and occluded facial expression images. Three typical facial features, i.e., the raw pixels, Gabor wavelets representation and local binary patterns (LBP), are extracted to evaluate the performance of the SRC method. Compared with the nearest neighbor (NN), linear support vector machines (SVM) and the nearest subspace (NS), experimental results on the popular Cohn-Kanade facial expression database demonstrate that the SRC method obtains better performance and stronger robustness to corruption and occlusion on robust facial expression recognition tasks.

No MeSH data available.


Related in: MedlinePlus