Papers by Keyword: Face Recognition

Paper TitlePage

Authors: Ti Jian Cai, Xiao Ping Fan, Jun Xu
Abstract: Empirical evidence shows that introducing additional structured priors can reduce complexity of coding data, and achieve better performance. To improve the performance of sparse representation-based classification (SRC), the article based on the potential correlations between the elements of dictionary gets a mixed group sparsity which is composed of dynamic group sparsity and fixed-length group sparsity. To solve the structured sparsity efficiently, structured greedy algorithm (structOMP) is redesigned to fit the new structure. The modification includes search space and its neighbor. Finally, three sparse models are compared by experiments of face recognition, and the results show that the mixed group sparsity can improve the face recognition rate of other sparse models by 10% or more in dealing with corrupted data.
Authors: Jai Yi Zhu, Yun Juan Liang, Li Li
Abstract: In this paper a face location technology based on gradient distributions is presented. It begins with vertical location by use of vertical integral projection on the two-valued image of the original image, and then proceeds with horizontal location according to the distributions of gradient direction. Experiments have proved this technology fast and efficient.
Authors: Yu Sun, Zhong Hui Lin, Ru Bo Zhang
Abstract: Data manifold represented in the reality is intrinsically in tensor form and so tensor-based subspace algorithms can preserve the intrinsic spatial structure information. They are beneficial for data representation and classification and have been widely used in recent years. In this paper, a new algorithm called Tensor based Isometric Projection (TIsoProjection) is proposed. The proposed algorithm can naturally describe the spatial relationship between the column vectors and the row vectors. Also it solves the small sample size (SSS) problem. Experiments on the ORL and YaleB demonstrate that the proposed algorithm can achieve higher recognition rate.
Authors: Yuan Xing Lv, Yan Ni Deng, Yuan Shi, Qiang Li, Wen Peng
Abstract: This paper proposes an adaptive discriminant linear local tangent space alignment algorithm DALLTSA. On the basis of LLTSA algorithm adding adaptive and discriminant gets DALLTSA.DALLTSA not only combines characteristics in DLLTSA that maintain the local geometry and meets the maximum between-class scatter matrix, but also dynamically selects K-neighbor better to reflect the degree of polymerization between samples. Finally, the face recognition experiments based on Gabor [1] filter and DALLTSA shows that this algorithm improves the recognition rate and robustness.
Authors: Xiao Kang Wu, Cheng Gang Xie, Qin Lu
Abstract: ion generation based on face detection and recognitionXiaokang Wu1, a, Chenggang Xie2, b, Qin Lu2, c1 National University Of Defense Technology, Changsha 410073, China;,,Keywords: face detection, face recognition, key frame, video abstractionAbstract. In order to facilitate users browse the behaviors and expressions of interesting objects in a video quickly, need to remove the redundancy information and extract key frames related to the object interested. This paper uses a fast face detection based on skin color, and recognition technology using spectrum feature matching, decompose the coupling video, and classify frames related to the object into different sets, generate a different video abstraction of each object. Experimental results show that the algorithm under different light conditions has better practicability.
Authors: Shi Ping Li, Yu Cheng, Hui Bin Liu, Lin Mu
Abstract: Linear Discriminant Analysis (LDA) [1] is a well-known method for face recognition in feature extraction and dimension reduction. To solve the “small sample” effect of LDA, Two-Dimensional Linear Discriminant Analysis (2DLDA) [2] has been used for face recognition recently,but its could hardly take use of the relationship between the adjacent scatter matrix. In this paper, I improved the between-class scatter matrix, proposed paired-class scatter matrix for face representation and recognition. In this new method, a paired between-class scatter matrix distance metric is used to measure the distance between random paired between-class scatter matrix. To test this new method, ORL face database is used and the results show that the paired between-class scatter matrix based 2DLDA method (N2DLDA) outperforms the 2DLDA method and achieves higher classification accuracy than the 2DLDA algorithm.
Authors: Ze Hua Zhou
Abstract: Recently, automatic face recognition method has become one of the key issues in the field of pattern recognition and artificial intelligence. Typically, the face recognition process can be divided into three parts: the detection and recognition of human face, facial feature extraction and face recognition, and among which the facial feature extraction is the key to face recognition technology. In this paper, an extraction algorithm of face recognition feature, which is based on face recognition feature, is proposed. The experimental results based on the ORL face database demonstrate that this algorithm works well.
Authors: Huai Ming Yang, Jin Guang Sun
Abstract: A new face image feature extraction and recognition algorithm based on Scale Invariant Feature Transform (SIFT) and Local Linary Patterns (LBP) is proposed in this paper. Firstly, a set of keypoints are extracted from images by using the SIFT algorithm; Secondly, each keypoint is described by LBP patterns; Finally, a combination of the global and local similarity is adopted to calculate the matching results for face images. Calculation results show that the algorithm can reduce the matching dimension of feature points, improve the recognition rate and perspective; it has nice robustness against the interferences such as rotation, lighting and expression.
Authors: Fang Min Hu, Hui Ya Zhao
Abstract: The feature extraction is a great important step for face recognition. When all features are extracted and selected for face recognition, it results in poor recognition rate because there are too many irrelevant, redundant and noisy features which also increase the time consumption. Therefore, a good feature selection method is necessary. This problem can be regarded as a combinatorial optimization solution. To overcome this problem, An improved kernel principal component analysis based on chaotic artificial fish school algorithm is proposed. The feature subspace of face pictures is obtained by standard kernel principal component analysis where a better feature subspace is selected by improved chaotic artificial fish school algorithm which based on couple chaotic maps increases the diversity of fish, has better global convergence ability and is not easy to fall into local optimum when facing with complex problems. The experimental results show that the proposed method has significantly improved the performance of conventional kernel principal component analysis.
Authors: Jin Qing Liu, Qun Zhen Fan, Dong Cao
Abstract: With the safety awareness strengthened, identification authentication technology has been increasingly concerned. Face recognition is attractive in pattern recognition and artificial intelligence field, and face feature extraction is a very important part in face recognition. This paper first introduced preprocessing of face images, PCA and ICA algorithm. Considering PCA and ICA their respective strengths and weaknesses, then a novel face feature extraction method based on PCA and ICA is proposed. The NN classifier is select to face classification and recognition on the ORL face database. From the actual requirements, the paper analyses hardware platforms based on DM642, and finally use tool CCS software to optimize program and implementation base on DM642 to meet the real-time requirements. Experiments indicated that the modified method is superior to PCA and ICA algorithm.
Showing 21 to 30 of 140 Paper Titles