TL;DR: Nonnegative Double Singular Value Decomposition (NNDSVD), a new method designed to enhance the initialization stage of nonnegative matrix factorization (NMF), is described and suggests that NNDSVD leads to rapid reduction of the approximation error of many NMF algorithms.

TL;DR: This paper considers a general problem of learning from pairwise constraints in the form of must-links and cannot-links, and aims to learn a Mahalanobis distance metric.

TL;DR: Experimental results show that the proposed robust path-based spectral clustering method consistently outperforms other methods due to its higher robustness, and comparisons with some other methods show this method to be significantly more robust than spectral clusters and path- based clustering.

TL;DR: This work proposes four new dynamic selection schemes which explore the properties of the oracle concept and suggests that the proposed schemes, using the majority voting rule for combining classifiers, perform better than the static selection method.

TL;DR: An improved voting scheme for the Hough transform is presented that allows a software implementation to achieve real-time performance even on relatively large images and produces a much cleaner voting map and makes the transform more robust to the detection of spurious lines.

TL;DR: Experiments with synthetic and real data sets show that the proposed ECM (evidential c-means) algorithm can be considered as a promising tool in the field of exploratory statistics.

TL;DR: The modified finite Radon transform is proposed, which can extract principal lines effectively and efficiently even in the case that the palmprint images contain many long and strong wrinkles.

TL;DR: It is shown that imputation with the tested methods on average improves classification accuracy when compared to classification without imputation, and some classifiers such as C4.5 and Nai@?ve-Bayes were found to be missing data resistant, i.e., they can produce accurate classification in the presence of missing data.

TL;DR: This paper presents a new model of camera lens distortion, according to which lens distortion is governed by the coefficients of radial distortion and a transform from ideal image plane to real sensor array plane.

TL;DR: By using the proposed algorithm, a 90% tampered image can be recovered to a dim yet still recognizable condition (PSNR ~20dB).

TL;DR: A novel minutiae-based fingerprint matching algorithm that ranks 1st on DB3, the most difficult database in FVC2002, and on the average ranks 2nd on all 4 databases.

TL;DR: This paper investigates the cospectrality of the various matrix representations over large graph and tree sets, extending the work of previous authors and shows that the Euclidean distance between spectra tracks the edit distance between graphs over a wide range of edit costs.

TL;DR: A novel secret image sharing scheme combining steganography and authentication based on Chinese remainder theorem (CRT) is proposed that not only improves the authentication ability but also enhances the visual quality of the stego images.

TL;DR: Experimental results show that the proposed ''Laplacianpalm'' approach provides a better representation and achieves lower error rates in palm recognition, and the proposed multimodal method outperforms any of its individual modality.

TL;DR: The quantised random projection ensemble based on the Johnson-Lindenstrauss Lemma is used to establish the mathematical foundation of BioHash, a form of cancellable biometrics which mixes a set of user-specific random vectors with biometric features and elucidate the characteristics of Bio hash in pattern recognition as well as security view points and propose new methods to rectify the stolen-token problem.

TL;DR: This paper proposes to use another form of supervision information for feature selection, i.e. pairwise constraints, which specifies whether a pair of data samples belong to the same class (must-link constraints) or different classes (cannot- link constraints).

TL;DR: This tutorial examines the problem area, its methods, successes and failures, focusing on the nature of the speech signal and techniques to accomplish useful data reduction, and compares it with other areas of PR.

TL;DR: A new version of the global k-means algorithm, an incremental algorithm that dynamically adds one cluster center at a time and uses each data point as a candidate for the k-th cluster center, is proposed.

TL;DR: A novel algorithm for image feature extraction, namely, the two-dimensional locality preserving projections (2DLPP), which directly extracts the proper features from image matrices based on locality preserving criterion is proposed.

TL;DR: This paper proposes a semi-supervised dimensionality reduction framework, which can efficiently handle the unlabeled data and can significantly improve the accuracy rates of the corresponding supervised and unsupervised approaches.

TL;DR: An integrated image fusion and match score fusion of multispectral face images using [email protected] SVM and Dezert Smarandache theory of fusion which is based on plausible and paradoxical reasoning is presented.

TL;DR: Zhang et al. as mentioned in this paper put forward a new method of co-occurrence matrix to describe image features, which can express the spatial correlation of textons, and quantized the original images into 256 colors and computed color gradient from the RGB vector space.

TL;DR: The suggested algorithm is based on several steps and has been tested on a large set of shape databases providing performances for both in recognition and in retrieval superior to most of previously proposed approaches.

TL;DR: The comparisons performed on a simple benchmark and on a ground truth image database show that with AFCC the results of clustering can be significantly improved with few constraints, making this semi-supervised approach an attractive alternative in the categorization of image databases.

TL;DR: The experimental results show that the proposed method performs well for face recognition problems, compared with conventional methods such as the principal component analysis (PCA), Fisher's linear discriminant (FLD), etc.

TL;DR: A model of memetic algorithm is proposed that incorporates an ad hoc local search specifically designed for optimizing the properties of prototype selection problem with the aim of tackling the scaling up problem.

TL;DR: A region-based image retrieval system with high-level semantic learning that supports both query by keyword and query by region of interest and outperforms two well-established decision tree induction algorithms ID3 and C4.5 in image semantic learning.

TL;DR: An overview on the last 40-years of technical advances in the field of character and document recognition in Japan is presented, and robustness design principles, which have proven to be effective to solve complex problems in postal address recognition are discussed.

TL;DR: The proposed scheme makes the number of secret images not restricted and further extends it to be general as a result, the proposed scheme enhances visual secret sharing schemes' ability for multiple secrets.