Local binary patterns for multi-view facial expression recognition

Moore, Stephen; Bowden, Richard

doi:10.1016/j.cviu.2010.12.001

Cited by 321 publications

(180 citation statements)

References 22 publications

Supporting

Mentioning

176

Contrasting

Order By: Relevance

“…Zernike moments are rotation invariant features, which can be used to address in-plane head pose variation. In the field of facial expression recognition, rotation invariant LBP and uniform LBP [11] have also been used to overcome the rotation problem. In [12], Quantised Local Zernike Moment (QLZM) is used to describe the neighbourhood of a face sub-region.…”

Section: Introductionmentioning

confidence: 99%

A dynamic framework based on local Zernike moment and motion history image for facial expression recognition

Fan

Tjahjadi

2017

Pattern Recognition

View full text Add to dashboard Cite

A dynamic descriptor facilitates robust recognition of facial expressions in video sequences. The current two main approaches to the recognition are basic emotion recognition and recognition based on facial action coding system (FACS) action units. In this paper we focus on basic emotion recognition and propose a spatiotemporal feature based on local Zernike moment in the spatial domain using motion change frequency. We also design a dynamic feature comprising motion history image and entropy. To recognise a facial expression, a weighting strategy based on the latter feature and sub-division of the image frame is applied to the former to enhance the dynamic information of facial expression, and followed by the application of the classical support vector machine. Experiments on the CK+ and MMI datasets using leave-one-out cross validation scheme demonstrate that the integrated framework achieves a better performance than using individual descriptor separately. Compared with six state-of-arts methods, the proposed framework demonstrates a superior performance.

show abstract

Section: Introductionmentioning

confidence: 99%

A dynamic framework based on local Zernike moment and motion history image for facial expression recognition

Fan

Tjahjadi

2017

Pattern Recognition

View full text Add to dashboard Cite

show abstract

“…From Table 3(a), we observe that VC-GPM outperforms, on average, the appearance-based methods. This difference is in part due to the features used and in part due to the fact that the methods in [3] and [9] both fail to model correlations between different views. By contrast, the CGP method accounts for the relations between the views in a pair-wise manner, while VC-GPM does so for all the views simultaneously.…”

Section: Comparisons With Other Multi-view Methodsmentioning

confidence: 94%

“…The results for the LGBP-based method are obtained from [3]. For the method in [9], we extracted Sparse SIFT (SSIFT) features from the same images we used from MultiPIE.…”

Section: Comparisons With Other Multi-view Methodsmentioning

confidence: 99%

“…Recent advances in the field, focus on view-invariant FER, and based on how they deal with the head-pose variations in 2D, they can be classified in those that: (i) perform view-invariant FER ( [3][4][5]), (ii) perform view normalization before FER ( [6,7]), and (iii) learn a single classifier using data from multiple views ( [8,9]). A representative of the first group is [3], where the Local Binary Patterns (LBP) [10] (and its variants) are used to perform a two-step FER: first, the view is classified using the Support Vectors Machine (SVM) [11], and then, a view-specific SVM is applied to perform FER.…”

Section: Introductionmentioning

confidence: 99%

“…A representative of the first group is [3], where the Local Binary Patterns (LBP) [10] (and its variants) are used to perform a two-step FER: first, the view is classified using the Support Vectors Machine (SVM) [11], and then, a view-specific SVM is applied to perform FER. In [4], different appearance features are extracted around the locations of characteristic facial points, and used to train various pose-specific classifiers.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

View-Constrained Latent Variable Model for Multi-view Facial Expression Classification

Eleftheriadis

Rudovic

Pantić

2014

Advances in Visual Computing

View full text Add to dashboard Cite

Abstract. We propose a view-constrained latent variable model for multi-view facial expression classification. In this model, we first learn a discriminative manifold shared by multiple views of facial expressions, followed by the expression classification in the shared manifold. For learning, we use the expression data from multiple views, however, the inference is performed using the data from a single view. Our experiments on data of posed and spontaneously displayed facial expressions show that the proposed approach outperforms the state-of-the-art methods for multi-view facial expression classification, and several state-of-theart methods for multi-view learning.

show abstract