Search

Traditional facial expression recognition methods usually assume that the training set and the test set are independent and identically distributed. However, in actual expression recognition applications, the conditions of independent and identical distribution are hardly satisfied for the training set and test set because of the difference of light, shade, race and so on. In order to solve this problem and improve the performance of expression recognition in the actual applications, a novel method based on transfer learning and sparse coding is applied to facial expression recognition. First of all, a common primitive model, that is, the dictionary is learnt. Then, based on the idea of transfer learning, the learned primitive pattern is transferred to facial expression and the corresponding feature representation is obtained by sparse coding. The experimental results in CK +, JAFFE and NVIE database shows that the transfer learning based on sparse coding method can effectively improve the expression recognition rate in the cross-domain expression recognition task and is suitable for the practical facial expression recognition applications. [ABSTRACT FROM AUTHOR]

A novel low-cost highly discriminatory feature space is introduced for facial expression recognition, which incorporates weight with the Optimized Local Directional Pattern (OLDP), capable of robust performance over a range of image resolutions. In addition, we use Adaboost to pick a small set of high-flying features, which are used by the Support Vector Machine (SVM) to classify facial expressions proficiently. Experimental results show that the proposed technique improves both the accuracy and the speed of the final classifier compares to other existing state-of-the-art methods. [ABSTRACT FROM PUBLISHER]

Both Gabor filters and Support Vector Machines (SVMs) are widely used in computer vision tasks for feature extraction and classification respectively. However the method is usually plagued by the problems of high computational complexity and memory usage owing to the high dimensionality of the Gabor filter responses. There were methods proposed to mitigate this problem by truncating or finding a gist of the responses but such approaches also lead to loss of information. Ashraf et al. gave a reinterpretation of the whole method and proposed a way to eliminate the need for such approximations. But they only give an analysis for linear SVM. This paper extends their work and provides analysis for nonlinear kernels within the same framework. The class of nonlinear kernels that are compatible with this framework are derived and experimental results on the facial expression recognition task are reported. [ABSTRACT FROM PUBLISHER]

The Internet of things uses in various fields, which is of great convenience to people's lives. Health monitoring can provide users with convenient and reliable health services, which are of low cost. The current health monitoring systems depend on the instruments. Based on the Internet of things, a system called NCHMS is proposed, which is monitoring the health status with no contact. This article elaborates on the structures and functions of NCHMS, improves the expression classification space for health monitoring, and gives a detailed expression recognition algorithm. The experimental data from the simulated experiments can tell that the system have considerable feasibility and practicality. [ABSTRACT FROM PUBLISHER]

A novel facial expression recognition approach based on 2DLDA and PSO is proposed in this paper. The multiresolution and multidirectional Contourlet transform is first adopted to decrease the data of expression images and extract frequency and direction elements which are beneficial to FER. Then the 2DLDA is used to extract expression features in order to reduce the dimensions of feature space. Finally, the PSO algorithm is used to optimize the features obtained. Moreover, the impact of iterations of PSO on recognition rate is researched based on the JAFEE database. Experimental results show that the method proposed in this paper is more robust and obtains higher accuracy. [ABSTRACT FROM PUBLISHER]

Given that sign language is used as a primary means of communication by as many as two million deaf individuals in the U.S. and as augmentative communication by hearing individuals with a variety of disabilities, the development of robust, real-time sign language recognition technologies would be a major step forward in making computers equally accessible to everyone. However, most research in the field of sign language recognition has focused on the manual component of signs, despite the fact that there is critical grammatical information expressed through facial expressions and head gestures. We propose a novel framework for robust tracking and analysis of facial expression and head gestures, with an application to sign language recognition. We then apply it to recognition with excellent accuracy (≥=95%) of two classes of grammatical expressions, namely wh-questions and negative expressions. Our method is signer-independent and builds on the popular "bag-of-words" model, utilizing spatial pyramids to model facial appearance and temporal pyramids to represent patterns of head pose changes. [ABSTRACT FROM AUTHOR]

Given that sign language is used as a primary means of communication by as many as two million deaf individuals in the U.S. and as augmentative communication by hearing individuals with a variety of disabilities, the development of robust, real-time sign language recognition technologies would be a major step forward in making computers equally accessible to everyone. However, most research in the field of sign language recognition has focused on the manual component of signs, despite the fact that there is critical grammatical information expressed through facial expressions and head gestures. We propose a novel framework for robust tracking and analysis of facial expression and head gestures, with an application to sign language recognition. We then apply it to recognition with excellent accuracy (≥=95%) of two classes of grammatical expressions, namely wh-questions and negative expressions. Our method is signer-independent and builds on the popular "bag-of-words" model, utilizing spatial pyramids to model facial appearance and temporal pyramids to represent patterns of head pose changes. [ABSTRACT FROM AUTHOR]

Conversation between two individuals requires verbal dialogue; the majority of human communication however consists of non-verbal cues such as gestures and facial expressions. Blind individuals are thus hindered in their interaction capabilities. To address this, we are building a computer vision system with facial recognition and expression algorithms to relay nonverbal messages to a blind user. The device will communicate the identities and facial expressions of communication partners in realtime. In order to ensure that this device will be useful to the blind community, we conducted surveys and interviews and we are working with subjects to test prototypes of the device. This paper describes the algorithms and design concepts incorporated in this device, and it provides a commentary on early survey and interview results. A corresponding poster with demonstration stills is exhibited at this conference. [ABSTRACT FROM AUTHOR]