In this paper, a new dynamic facial expression recognition method is proposed. Dynamic facial expression recognition is formulated as a longitudinal groupwise registration problem. The main contributions of this method lie in the following aspects: 1) subject-specific facial feature movements of different expressions are described by a diffeomorphic growth model; 2) salient longitudinal facial expression atlas is built for each expression by a sparse groupwise image registration method, which can describe the overall facial feature changes among the whole population and can suppress the bias due to large intersubject facial variations; and 3) both the image appearance information in spatial domain and topological evolution information in temporal domain are used to guide recognition by a sparse representation method. The proposed framework has been extensively evaluated on five databases for different applications: the extended Cohn-Kanade, MMI, FERA, and AFEW databases for dynamic facial expression recognition, and UNBC-McMaster database for spontaneous pain expression monitoring. This framework is also compared with several state-of-the-art dynamic facial expression recognition methods. The experimental results demonstrate that the recognition rates of the new method are consistently higher than other methods under comparison.

Video texture synthesis is the process of providing a continuous and infinitely varying stream of frames, which plays an important role in computer vision and graphics. However, it still remains a challenging problem to generate high-quality synthesis results. Considering the two key factors that affect the synthesis performance, frame representation and blending artifacts, we improve the synthesis performance from two aspects: 1) Effective frame representation is designed to capture both the image appearance information in spatial domain and the longitudinal information in temporal domain. 2) Artifacts that degrade the synthesis quality are significantly suppressed on the basis of a diffeomorphic growth model. The proposed video texture synthesis approach has two major stages: video stitching stage and transition smoothing stage. In the first stage, a video texture synthesis model is proposed to generate an infinite video flow. To find similar frames for stitching video clips, we present a new spatial-temporal descriptor to provide an effective representation for different types of dynamic textures. In the second stage, a smoothing method is proposed to improve synthesis quality, especially in the aspect of temporal continuity. It aims to establish a diffeomorphic growth model to emulate local dynamics around stitched frames. The proposed approach is thoroughly tested on public databases and videos from the Internet, and is evaluated in both qualitative and quantitative ways.

An image-based visual speech animation system is presented in this paper. A video model is proposed to preserve the video dynamics of a talking face. The model represents a video sequence by a low-dimensional continuous curve embedded in a path graph and establishes a map from the curve to the image domain. When selecting video segments for synthesis, we loosen the traditional requirement of using triphone as the unit to allow segments to contain longer natural talking motion. Dense videos are sampled from the segments, concatenated, and downsampled to train a video model that enables efficient time alignment and motion smoothing for the final video synthesis. Different viseme definitions are used to investigate the impact of visemes on the video realism of the animated talking face. The system is built on a public database and tested both objectively and subjectively.

We investigate rotation invariant image description and develop a linear model based descriptor namely MiC, which is suited to modeling microscopic configuration of images. To explore multi-channel discriminative information of both the microscopic configuration and local structures, the feature extraction process is formulated as an unsupervised framework that consists of: 1) the configuration model to encode image microscopic configuration; and 2) local patterns to describe local structural information. In this way, images are represented by a novel feature: local configuration pattern (LCP). We evaluate the performance of the proposed method by classifying textures present in three challenging texture databases: Outex_TC_00012, KTH-TIPS2 and Columbia-Utrecht (CUReT). The encouraging results show that LCPs is highly discriminative. 1

Dynamic textures are image sequences with visual pattern repetition in time and space, such as smoke, flames, moving objects and so on. Dynamic texture synthesis is to provide a continuous and infinitely varying stream of images by doing operations on dynamic textures. Considering that the previous video texture method provides high-quality visual results, but its representation does not well explore the temporal correlation among frames, we develop a novel spatial temporal descriptor for frame description accompanied with a similarity measure on the basis of the video texture method. Compared with the previous one, our method considers both the spatial and temporal domains of video sequences in representation; moreover, combines the local and global description on each spatial-temporal plane. From experimental results, the proposed method achieves better performance in both the syntheses of natural scene and human motion. Especially, it has the characteristic to be robust to noise in remodeling videos into infinite time domain.

A new ear recognition approach, including a feature extraction method and the recognition framework, is presented in this paper. The proposed feature extraction method, called the Local Similarity Binary Pattern (LSBP), considers both the connectivity and similarity information in representation. In ear recognition, LSBP is combined with the Local Binary Pattern (LBP) to represent the ear image. The concatenated histogram sequences encode more relationships among neighborhoods that are shown to be discriminative. To enhance efficient representation, Cellular Neural Network is adopted to preprocess images, the function of which is to eliminate irrelevant information. From the experimental results conducted on the USTB ear database, the proposed approach outperforms some other well-known methods in terms of the recognition rate.

A new image representation method is proposed for face recognition in this paper, called local Gabor phase difference pattern (LGPDP). Unlike the Histogram of Gabor Phase Patterns (HGPP) that exploits the relationships of Gabor phase between neighborhood pixels, the LGPDP captures the Gabor phase difference relationships to represent the image. In order to avoid the sensitivity of the Gabor phase to location variations, the feature encodes the discriminative information in an elaborate way. The impressive experimental results of the proposed method compared with some other state-of-art methods conducted on the FERET database and the ORL face database demonstrate its efficiency and validity.

As face is a topological object, spatial contents contained in facial images (i.e. eyes, nose…) play an important role in feature extraction. To preserve spatial information, region decomposition is an essential step in face recognition for local feature based methods. In this paper, a new region decomposition method is proposed based on Cellular Neural Network (CNN). This method, called Face Penta-Chotomy (FPC), can be factorized into two parts. First, a stable facial region is extracted by a CNN template. Then other four regions are depicted according to the stable facial region and facial proportion. The local binary pattern (LBP) is adopted as the region descriptor. This method is evaluated by conducting experiments on the Yale face database B and ORL database. Besides, it compared with six state-of-the-art methods. From experimental results, it outperforms all the compared methods and the feature dimension can be significantly reduced compared with the conventional uniform region decomposition method. Moreover, the proposed method is demonstrated to be robust under single training condition.

A number of templates for image translation using cellular neural network (CNN) have been proposed before. In this paper, all cases of the 3 × 3 uncoupled CNN template for translation of gray-scale images are investigated and their functions are discussed.