Type-2 fuzzy Gaussian mixture models

This paper presents a new extension of Gaussian mixture models (GMMs) based on type-2 fuzzy sets (T2 FSs) referred to as T2 FGMMs. The estimated parameters of the GMM may not accurately reflect the underlying distributions of the observations because of insufficient and noisy data in real-world problems. By three-dimensional membership functions of T2 FSs, T2 FGMMs use footprint of uncertainty (FOU) as well as interval secondary membership functions to handle GMMs uncertain mean vector or uncertain covariance matrix, and thus GMMs parameters vary anywhere in an interval with uniform possibilities. As a result, the likelihood of the T2 FGMM becomes an interval rather than a precise real number to account for GMMs uncertainty. These interval likelihoods are then processed by the generalized linear model (GLM) for classification decision-making. In this paper we focus on the role of the FOU in pattern classification. Multi-category classification on different data sets from UCI repository shows that T2 FGMMs are consistently as good as or better than GMMs in case of insufficient training data, and are also insensitive to different areas of the FOU. Based on T2 FGMMs, we extend hidden Markov models (HMMs) to type-2 fuzzy HMMs (T2 FHMMs). Phoneme classification in the babble noise shows that T2 FHMMs outperform classical HMMs in terms of the robustness and classification rate. We also find that the larger area of the FOU in T2 FHMMs with uncertain mean vectors performs better in classification when the signal-to-noise ratio is lower.     

Piecewise affine kernel tracking for non-planar targets

Since the multiple kernel representation opened in tracking the possibility of representing several features of the target in the same model, tracking multiple features using kernel-based methods has received a great attention. In spite of these efforts, the formulation has been reduced to tracking planar targets or targets rotating inside a plane parallel to the image plane. The aim of this paper is to extend the multi-kernel tracking to cope with situations different to those. To this end, we consider the triangular mesh described by the centers of the kernels and we develop the estimation of a set of affine transforms, one at each mesh triangle, subject to the constraints that each affine transform of a triangle must be compatible with the affine transforms coming from contiguous triangles. The method is applied to sequences including face and car tracking. Results show an outperformance respect to previous kernel tracking methods, which generally work with a too restricted set of movements.     

Shape from shading using graph cuts

In this paper, we investigate the applicability of graph cuts to the SFS (shape-from-shading) problem. We propose a new semi-global method for SFS using graph cuts. The new algorithm combines the local method proposed by Lee and Rosenfeld [C.H. Lee, A. Rosenfeld, Improved methods of estimating shape from shading using the light source coordinate system, Artif. Intell. 26 (1985) 125-143] and a global method using an energy minimization technique. By employing a new global energy minimization formulation, the convex/concave ambiguity problem of Lee and Rosenfeld's method can be resolved efficiently. A new combinatorial optimization technique, the graph cuts method, is used for the minimization of the proposed energy functional. Experimental results on a variety of synthetic and real-world images show that the proposed algorithm reconstructs the 3-D shape of objects very efficiently.     

Face recognition using adaptively weighted patch PZM array from a single exemplar image per person

Though numerous approaches have been proposed for face recognition, little attention is given to the moment-based face recognition techniques. In this paper we propose a novel face recognition approach based on adaptively weighted patch pseudo Zernike moment array (AWPPZMA) when only one exemplar image per person is available. In this approach, a face image is represented as an array of patch pseudo Zernike moments (PPZM) extracted from a partitioned face image containing moment information of local areas instead of global information of a face. An adaptively weighting scheme is used to assign proper weights to each PPZM to adjust the contribution of each local area of a face in terms of the quantity of identity information that a patch contains and the likelihood of a patch is occluded. An extensive experimental investigation is conducted using AR and Yale face databases covering face recognition under controlled/ideal conditions, different illumination conditions, different facial expressions and partial occlusion. The system performance is compared with the performance of four benchmark approaches. The encouraging experimental results demonstrate that moments can be used for face recognition and patch-based moment array provides a novel way for face representation and recognition in single model databases.     

Object detection by global contour shape

We present a method for object class detection in images based on global shape. A distance measure for elastic shape matching is derived, which is invariant to scale and rotation, and robust against non-parametric deformations. Starting from an over-segmentation of the image, the space of potential object boundaries is explored to find boundaries, which have high similarity with the shape template of the object class to be detected. An extensive experimental evaluation is presented. The approach achieves a remarkable detection rate of 83-91% at 0.2 false positives per image on three challenging data sets.     

Locally linear discriminant embedding: An efficient method for face recognition

In this paper an efficient feature extraction method named as locally linear discriminant embedding (LLDE) is proposed for face recognition. It is well known that a point can be linearly reconstructed by its neighbors and the reconstruction weights are under the sum-to-one constraint in the classical locally linear embedding (LLE). So the constrained weights obey an important symmetry: for any particular data point, they are invariant to rotations, rescalings and translations. The latter two are introduced to the proposed method to strengthen the classification ability of the original LLE. The data with different class labels are translated by the corresponding vectors and those belonging to the same class are translated by the same vector. In order to cluster the data with the same label closer, they are also rescaled to some extent. So after translation and rescaling, the discriminability of the data will be improved significantly. The proposed method is compared with some related feature extraction methods such as maximum margin criterion (MMC), as well as other supervised manifold learning-based approaches, for example ensemble unified LLE and linear discriminant analysis (En-ULLELDA), locally linear discriminant analysis (LLDA). Experimental results on Yale and CMU PIE face databases convince us that the proposed method provides a better representation of the class information and obtains much higher recognition accuracies.     

1D correlation filter based class-dependence feature analysis for face recognition

In this paper, a novel one-dimensional correlation filter based class-dependence feature analysis (1D-CFA) method is presented for robust face recognition. Compared with original CFA that works in the two dimensional (2D) image space, 1D-CFA encodes the image data as vectors. In 1D-CFA, a new correlation filter called optimal extra-class origin output tradeoff filter (OEOTF), which is designed in the low-dimensional principal component analysis (PCA) subspace, is proposed for effective feature extraction. Experimental results on benchmark face databases, such as FERET, AR, and FRGC, show that OEOTF based 1D-CFA consistently outperforms other state-of-the-art face recognition methods. This demonstrates the effectiveness and robustness of the novel method.     

Locally linear reconstruction for instance-based learning

Instance-based learning (IBL), so called memory-based reasoning (MBR), is a commonly used non-parametric learning algorithm. k-nearest neighbor (k-NN) learning is the most popular realization of IBL. Due to its usability and adaptability, k-NN has been successfully applied to a wide range of applications. However, in practice, one has to set important model parameters only empirically: the number of neighbors (k) and weights to those neighbors. In this paper, we propose structured ways to set these parameters, based on locally linear reconstruction (LLR). We then employed sequential minimal optimization (SMO) for solving quadratic programming step involved in LLR for classification to reduce the computational complexity. Experimental results from 11 classification and eight regression tasks were promising enough to merit further investigation: not only did LLR outperform the conventional weight allocation methods without much additional computational cost, but also LLR was found to be robust to the change of k.     

A superscalar simulation employing poisson distributed stalls

This paper presents a statistical approach to estimating the performance of a superscalar processor. Traditional trace-driven simulators can take a large amount time to conduct a performance evaluation of a machine, especially as the number of instructions increases. The result of this type of simulation is typically tied to the particular trace that was run. Elements such as dependencies, delays, and stalls are all a direct result of the particular trace being run, and can differ from trace to trace. This paper describes a model designed to separate simulation results from a specific trace. Rather than running a trace-driven simulation, a statistical model is employed, more specifically a Poisson distribution, to predict how these types of delay affects performance. Through the use of this statistical model, a performance evaluation can be conducted using a general code model, with specific stall rates, rather than a particular code trace. This model allows simulations to quickly run tens of millions of instructions and evaluate the performance of a particular micro-architecture while at the same time, allowing the flexibility to change the structure of the architecture.     

Processing and interactive editing of huge point clouds from 3D scanners

This paper describes a new out-of-core multi-resolution data structure for real-time visualization, interactive editing and externally efficient processing of large point clouds. We describe an editing system that makes use of the novel data structure to provide interactive editing and preprocessing tools for large scanner data sets. Using the new data structure, we provide a complete tool chain for 3D scanner data processing, from data preprocessing and filtering to manual touch-up and real-time visualization. In particular, we describe an out-of-core outlier removal and bilateral geometry filtering algorithm, a toolset for interactive selection, painting, transformation, and filtering of huge out-of-core point-cloud data sets and a real-time rendering algorithm, which all use the same data structure as storage backend. The interactive tools work in real-time for small model modifications. For large scale editing operations, we employ a two-resolution approach where editing is planned in real-time and executed in an externally efficient offline computation afterwards. We evaluate our implementation on example data sets of sizes up to 63 GB, demonstrating that the proposed technique can be used effectively in real-world applications.