We present local spline regression: a new approach to semi-supervised classi cation. The core idea of our approach is to introduce splines developed in Sobolev space to map the data points to be class labels. Speci cally, in each data neighborhood, an optimal spline is estimated under the regularized least squares regression framework. With this spline, the neighboring data points are mapped, and the quadratic loss is evaluated and then formulated in terms of class label vector. Such local losses evaluated on all of the neighborhoods are nally accumulated together to construct a learning model with global consistency. Finally, a transductive classi cation algorithm is developed. Comparative classification experiments on many public data sets and applications to interactive image segmentation and image matting illustrate the validity of our method.

Stability (robustness) of feature selection methods is a topic of recent interest yet often neglected importance with direct impact on the reliability of machine learning systems. We investigate the problem of evaluating the stability of feature selection processes yielding subsets of varying size. We introduce several novel feature selection stability measures and adjust some existing measures in a unifying framework that offers broad insight into the stability problem. We study in detail the properties of considered measures and demonstrate on various examples what information about the feature selection process can be gained. We also introduce an alternative approach to feature selection evaluation in form of measures that enable comparing the similarity of two feature selection processes. These measures enable comparing, e.g., the output of two feature selection methods or two runs of one method with different parameters. The information obtained using the considered stability and similarity measures is shown usable for assessing feature selection methods (or criteria) as such.

There is now a growing need to identify various kinds of activities that occur in videos. In this paper, we first present a logical language called Probabilistic Activity Description Language (PADL) in which users can specify activities of interest. We then develop a probabilistic framework which assigns to any subvideo of a given video sequence a probability that the subvideo contains the given activity, and we finally develop two fast algorithms to detect activities within this framework. OffPad finds all minimal segments of a video that contain a given activity with a probability exceeding a given threshold. In contrast, the OnPad algorithm examines a video during playout (rather than afterwards as OffPad does) and computes the probability that a given activity is occurring (even if the activity is only partially complete). Our prototype Probabilistic Activity Detection System (PADS) implements the framework and the two algorithms, building on top of existing image processing algorithms. We have conducted detailed experiments and compared our approach to four different approaches presented in the literature. We show that - for complex activity definitions - our approach outperforms all the other approaches.

Presented By: NEC
 
Ads by Pheedo

This paper addresses the problem of correspondence establishment in binocular stereo vision. We suggest a novel spatially continuous approach for stereo matching based on the variational framework. The proposed method suggests a unique regularization term based on Mumford-Shah functional for discontinuity preserving, combined with a new energy functional for occlusion handling. The evaluation process is based on concurrent minimization of two coupled energy functionals, one for domain segmentation (occluded vs. visible) and the other for disparity evaluation. In addition to a dense disparity map, our method also provides an estimation for the half-occlusion domain, and a discontinuity function allocating the disparity/depth boundaries. Two new constraints are introduced improving the revealed discontinuity map. The experimental tests include a wide range of real data sets from Middlebury stereo database. The results demonstrate the capability of our method in calculating an accurate disparity function with sharp discontinuities and occlusion map recovery. Significant improvements are shown comparing to a recently published variational stereo approach. A comparison on the Middlebury stereo benchmark with sub-pixel accuracies shows that our method is currently among the top-ranked stereo matching algorithms.

We propose a novel method for removing irrelevant frames from a video given user-provided frame-level labeling for a very small number of frames. We first hypothesize a number of windows which possibly contain the object of interest, and then figure out which window(s) truly contain the object of interest. Our method enjoys several favorable properties. First, compared to approaches where a single descriptor is used to describe a whole frame, each window's feature descriptor has the chance of genuinely describing the object of interest, hence it is less affected by background clutter. Second, by considering the temporal continuity of a video instead of treating frames as independent, we can hypothesize the location of the windows more accurately. Third, by infusing prior knowledge into the patch-level model, we can precisely follow the trajectory of the object of interest. This allows us to largely reduce the number of windows and hence reduce the chance of overfitting the data during learning. We demonstrate the effectiveness of the method by comparing it to several other semi-supervised learning approaches on challenging video clips.

A variety of different scripts are used in writing languages throughout the world. In a multi-script, multilingual environment, it is essential to know the script used in writing a document before an appropriate character recognition and document analysis algorithm can be chosen. In view of this, several methods for automatic script identification have been developed so far. They mainly belong to two broad categories – structure-based and visual appearance-based techniques. This survey report gives an overview of the different script identification methodologies under each of these categories. Methods for script identification in online data and video-texts are also presented. It is noted that the research in this field is relatively thin and still more research is to be done, particularly in case of handwritten documents.

This paper studies the training of support vector machine (SVM) classifiers with respect to the minimax and Neyman-Pearson criteria. In principle, these criteria can be optimized in a straightforward way using a cost-sensitive SVM. In practice, however, because these criteria require especially accurate error estimation, standard techniques for tuning SVM parameters, such as cross-validation, can lead to poor classifier performance. To address this issue, we first prove that the usual cost-sensitive SVM, here called the 2C-SVM, is equivalent to another formulation called the 2ν-SVM. We then exploit a characterization of the 2ν-SVM parameter space to develop a simple yet powerful approach to error estimation based on smoothing. In an extensive experimental study we demonstrate that smoothing significantly improves the accuracy of cross-validation error estimates, leading to dramatic performance gains. Furthermore, we propose coordinate descent strategies that offer significant gains in computational efficiency, with little to no loss in performance.

Presented By: NEC
 
Ads by Pheedo

In the appearance-based approach to face detection we generally learn the image statistics describing the texture pattern of the object class we want to detect in the image. Using this approach, algorithms now exist which can quite reliably detect faces in clutter in real time. However, these algorithms have had a limited success in providing an accurate and detailed description of the facial features. In general, this is due to the limited amount of information carried by the learned model. We resolve this problem by adding context information. This means that when we now search for a face or facial feature we look for those locations that most resemble the feature and those that are most dissimilar to its context. This task is, however, a difficult one, since the context and the texture of facial features vary widely under changing expression, pose and illumination. We address this problem with the use of subclass divisions. We divide the training samples of each facial feature into a set of subclasses, each representing a distinct construction of the same facial component (e.g., close versus open eyes) or its context (e.g., different hairstyle). We provide extensive experimental results on a total of 3,930 images.

This paper presents a video metrology approach using an uncalibrated single camera that is either stationary or in planar motion. Although theoretically simple, measuring the length of even a line segment in a given video is often a difficult problem. Most existing techniques for this task are extensions of single image-based techniques and do not achieve the desired accuracy especially in noisy environments. In contrast, the proposed algorithm moves line segments on the reference plane to share a common endpoint using the vanishing line information followed by fitting multiple concentric circles on the image plane. A fully automated real-time system based on this algorithm has been developed to measure vehicle wheelbases using an uncalibrated stationary camera. The system estimates the vanishing line using invariant lengths on the reference plane from multiple frames rather than the given parallel lines, which may not exist in videos. It is further extended to a camera undergoing a planar motion by automatically selecting frames with similar vanishing lines from the video. Experimental results show that the measurement results are accurate enough to classify moving vehicles based on their size.

Ma and Sonka proposed a fully parallel 3D thinning algorithm which does not always preserve topology. We propose an algorithm based on P-simple points which automatically corrects Ma and Sonka's Algorithm. As far as we know, our algorithm is the only fully parallel curve thinning algorithm which preserves topology.

This paper proposes a framework for single-image super-resolution. The underlying idea is to learn a map from input low-resolution images to target high-resolution images based on example pairs of input and output images. Kernel ridge regression (KRR) is adopted for this purpose. To reduce the time complexity of training and testing for KRR, a sparse solution is found by combining the ideas of kernel matching pursuit and gradient descent. As a regularized solution, KRR leads to a better generalization than simply storing the examples as it has been done in existing example-based algorithms and results in much less noisy images. However, this may introduce blurring and ringing artifacts around major edges as sharp changes are penalized severely. A prior model of a generic image class which takes into account the discontinuity property of images is adopted to resolve this problem. Comparison with existing algorithms shows the effectiveness of the proposed method.

This paper addresses the problem of self-validated labeling of Markov random fields (MRFs), namely to optimize an MRF with unknown number of labels. We present graduated graph cuts (GGC), a new technique that extends the binary s-t graph cut for self-validated labeling. Specifically, we use the split-and-merge strategy to decompose the complex problem to a series of tractable subproblems. In terms of Gibbs energy minimization, a suboptimal labeling is gradually obtained based upon a set of cluster-level operations. By using different optimization structures, we propose three practical algorithms: tree-structured graph cuts (TSGC), net-structured graph cuts (NSGC) and hierarchical graph cuts (HGC). In contrast to previous methods, the proposed algorithms can automatically determine the number of labels, properly balance the labeling accuracy, spatial coherence and the labeling cost (i.e., the number of labels), and are computationally efficient, independent to initialization and able to converge to good local minima. We apply the proposed algorithms to natural image segmentation. Experimental results show that our algorithms produce generally feasible segmentations for Benchmark datasets, and outperform alternative methods in terms of robustness to noise, speed and preservation of soft boundaries.

Multibody structure from motion (SfM) is the extension of classical SfM to dynamic scenes with multiple rigidly moving objects. Recent research has unveiled some of the mathematical foundations of the problem, but a practical algorithm, which can handle realistic sequences, is still missing. In this paper, we discuss the requirements for such an algorithm, highlight theoretical issues and practical problems, and describe how a static structure-from-motion framework needs to be extended to handle real dynamic scenes. Theoretical issues include different situations, in which the number of independently moving scene objects changes: moving objects can enter or leave the field of view, merge into the static background (e.g. when a car is parked), or split off the background and start moving independently. Practical issues arise due to small freely moving foreground objects with few and short feature tracks. We argue that all these difficulties need to be handled online, as structure-from-motion estimation progresses, and present an exemplary solution using the framework of probabilistic model-scoring.

We consider the problem of registering two observations on an arbitrary object, where the two are related by a geometric affine transformation of their coordinate systems, and by a non-linear mapping of their intensities. More generally, the framework is that of jointly estimating the geometric and radiometric deformations relating two observations on the same object. We show that the original high-dimensional, non-linear, non-convex search problem of simultaneously recovering the geometric and radiometric deformations can be represented by an equivalent sequence of two linear systems. A solution of this sequence yields an exact, explicit, and efficient solution to the joint estimation problem.

Choosing the appropriate parameter prior distributions associated to a given Bayesian model is a challenging problem. Conjugate priors can be selected for simplicity motivations. However conjugate priors can be too restrictive to model accurately the available prior information. This paper studies a new generative supervised classifier which assumes that the parameter prior distributions conditioned on each class are mixtures of Dirichlet processes. The motivations for using mixtures of Dirichlet processes is their known ability to model accurately a large class of probability distributions. A Monte carlo method allowing one to sample according to the resulting class conditional posterior distributions is then studied. The parameters appearing in the class conditional densities can then be estimated using these generated samples (following Bayesian learning). The proposed supervised classifier is applied to the classification of altimetric waveforms backscattered from different surfaces (oceans, ices, forests, deserts). This classification is a first step before developing tools allowing for the extraction of useful geophysical information from altimetric waveforms backscattered from non-oceanic surfaces.

The neighbourhood counting measure (NCM) is a similarity measure for multivariate data presented in an PAMI paper in 2006 by Wang. The comment paper by Argentini and Blanzieri refutes a remark in this paper about the minimum risk metric (MRM), and also compares MRM and NCM through experiments. This note is a response to the comment paper. The basis of the original remark is clarified by a combination of theoretical analysis of different implementations of MRM and experimental results using straightforward implementations of MRM and NCM.

Recently, there has been a growing interest in multi-way probabilistic clustering. Some efficient algorithms have been developed for this problem. However, no much attention has been paid on how to detect the number of clusters for the general n-way clustering (n>=2). To fill this gap, this problem is investigated based on n-way algebraic theory in this paper. A simple yet efficient detection method is proposed by eigenvalue decomposition (EVD), which is easy to implement. We justify this method. In addition, its effectiveness is demonstrated by the experiments on both simulated and real world data sets.

One of the challenges in automatic face recognition is to achieve temporal invariance. In other words, the goal is to come up with a representation and matching scheme that is robust to changes due to facial aging. Facial aging is a complex process that affects both the 3D shape of the face and its texture (e.g., wrinkles). These shape and texture changes degrade the performance of automatic face recognition systems. However, facial aging has not received substantial attention compared to other facial variations due to pose, lighting, and expression. We propose a 3D aging modeling technique and show how it can be used to compensate for the age variations to improve the face recognition performance. The aging modeling technique adapts view invariant 3D face models to the given 2D face aging database. The proposed approach is evaluated on three different databases (i.g., FG-NET, MORPH and BROWNS) using FaceVACS, a state-of-the-art commercial face recognition engine.

This paper presents a set of imperfectly symmetric measures based on a series of geometric transformation operations for quantitatively measuring the "amount" of symmetry of arbitrary shapes. The definitions of both bilateral symmetricity and rotational symmetricity give new insight into analyzing geometrical property of a shape and enable characterizing arbitrary shapes in a new way. We developed a set of criteria for quantitative galaxy classification using our proposed irregular shape symmetry measures. Our study has demonstrated effectiveness of the proposed method for the characterization of the shape of the celestial bodies. The concepts described in the paper are applicable to many fields such as mathematics, artificial intelligence, digital image processing, robotics, biomedicine etc.

We propose a shape-based, hierarchical part-template matching approach to simultaneous human detection and segmentation combining local part-based and global shape template-based schemes. The approach relies on the key idea of matching a part-template tree to images hierarchically to detect humans and estimate their poses. For learning a generic human detector, a pose-adaptive feature computation scheme is developed based on the tree matching approach. Instead of traditional concatenation-style image location-based feature encoding, we extract features adaptively in the context of human poses and train a kernel-SVM classifier to separate human/non-human patterns. Specifically, the features are collected in the local context of poses by tracing around the estimated shape boundaries. We also introduce an approach to multiple occluded human detection and segmentation based on an iterative occlusion compensation scheme. The output of our learned generic human detector can be used as an initial set of human hypotheses for the iterative optimization. We evaluate our approaches on three public pedestrian datasets (INRIA, MIT-CBCL, and USC-B) and two crowded sequences from Caviar Benchmark and Munich Airport datasets.