Showing papers on "Feature (computer vision) published in 2005"

Histograms of oriented gradients for human detection

Abstract: We study the question of feature sets for robust visual object recognition; adopting linear SVM based human detection as a test case. After reviewing existing edge and gradient based descriptors, we show experimentally that grids of histograms of oriented gradient (HOG) descriptors significantly outperform existing feature sets for human detection. We study the influence of each stage of the computation on performance, concluding that fine-scale gradients, fine orientation binning, relatively coarse spatial binning, and high-quality local contrast normalization in overlapping descriptor blocks are all important for good results. The new approach gives near-perfect separation on the original MIT pedestrian database, so we introduce a more challenging dataset containing over 1800 annotated human images with a large range of pose variations and backgrounds.

Toward integrating feature selection algorithms for classification and clustering

Abstract: This paper introduces concepts and algorithms of feature selection, surveys existing feature selection algorithms for classification and clustering, groups and compares different algorithms with a categorizing framework based on search strategies, evaluation criteria, and data mining tasks, reveals unattempted combinations, and provides guidelines in selecting feature selection algorithms. With the categorizing framework, we continue our efforts toward-building an integrated system for intelligent feature selection. A unifying platform is proposed as an intermediate step. An illustrative example is presented to show how existing feature selection algorithms can be integrated into a meta algorithm that can take advantage of individual algorithms. An added advantage of doing so is to help a user employ a suitable algorithm without knowing details of each algorithm. Some real-world applications are included to demonstrate the use of feature selection in data mining. We conclude this work by identifying trends and challenges of feature selection research and development.

Laplacian Score for Feature Selection

Abstract: In supervised learning scenarios, feature selection has been studied widely in the literature. Selecting features in unsupervised learning scenarios is a much harder problem, due to the absence of class labels that would guide the search for relevant information. And, almost all of previous unsupervised feature selection methods are "wrapper" techniques that require a learning algorithm to evaluate the candidate feature subsets. In this paper, we propose a "filter" method for feature selection which is independent of any learning algorithm. Our method can be performed in either supervised or unsupervised fashion. The proposed method is based on the observation that, in many real world classification problems, data from the same class are often close to each other. The importance of a feature is evaluated by its power of locality preserving, or, Laplacian Score. We compare our method with data variance (unsupervised) and Fisher score (supervised) on two data sets. Experimental results demonstrate the effectiveness and efficiency of our algorithm.

Feature models, grammars, and propositional formulas

Abstract: Feature models are used to specify members of a product-line. Despite years of progress, contemporary tools often provide limited support for feature constraints and offer little or no support for debugging feature models. We integrate prior results to connect feature models, grammars, and propositional formulas. This connection allows arbitrary propositional constraints to be defined among features and enables off-the-shelf satisfiability solvers to debug feature models. We also show how our ideas can generalize recent results on the staged configuration of feature models.

Infinite latent feature models and the Indian buffet process

Abstract: We define a probability distribution over equivalence classes of binary matrices with a finite number of rows and an unbounded number of columns. This distribution is suitable for use as a prior in probabilistic models that represent objects using a potentially infinite array of features. We identify a simple generative process that results in the same distribution over equivalence classes, which we call the Indian buffet process. We illustrate the use of this distribution as a prior in an infinite latent feature model, deriving a Markov chain Monte Carlo algorithm for inference in this model and applying the algorithm to an image dataset.

Robust global registration

Abstract: We present an algorithm for the automatic alignment of two 3D shapes (data and model), without any assumptions about their initial positions. The algorithm computes for each surface point a descriptor based on local geometry that is robust to noise. A small number of feature points are automatically picked from the data shape according to the uniqueness of the descriptor value at the point. For each feature point on the data, we use the descriptor values of the model to find potential corresponding points. We then develop a fast branch-and-bound algorithm based on distance matrix comparisons to select the optimal correspondence set and bring the two shapes into a coarse alignment. The result of our alignment algorithm is used as the initialization to ICP (iterative closest point) and its variants for fine registration of the data to the model. Our algorithm can be used for matching shapes that overlap only over parts of their extent, for building models from partial range scans, as well as for simple symmetry detection, and for matching shapes undergoing articulated motion.

Formalizing cardinality‐based feature models and their specialization

Abstract: Feature modeling is an important approach to capture the commonalities and variabilities in system families and product lines. Cardinality-based feature modeling integrates a number of existing extensions of the original feature-modeling notation from Feature-Oriented Domain Analysis. Staged configuration is a process that allows the incremental configuration of cardinality-based feature models. It can be achieved by performing a step-wise specialization of the feature model. In this article, we argue that cardinality-based feature models can be interpreted as a special class of context-free grammars. We make this precise by specifying a translation from a feature model into a context-free grammar. Consequently, we provide a semantic interpretation for cardinality-based feature models by assigning an appropriate semantics to the language recognized by the corresponding grammar. Finally, we give an account on how feature model specialization can be formalized as transformations on the grammar equivalent of feature models. Copyright © 2005 John Wiley & Sons, Ltd.

Feature bagging for outlier detection

Abstract: Outlier detection has recently become an important problem in many industrial and financial applications. In this paper, a novel feature bagging approach for detecting outliers in very large, high dimensional and noisy databases is proposed. It combines results from multiple outlier detection algorithms that are applied using different set of features. Every outlier detection algorithm uses a small subset of features that are randomly selected from the original feature set. As a result, each outlier detector identifies different outliers, and thus assigns to all data records outlier scores that correspond to their probability of being outliers. The outlier scores computed by the individual outlier detection algorithms are then combined in order to find the better quality outliers. Experiments performed on several synthetic and real life data sets show that the proposed methods for combining outputs from multiple outlier detection algorithms provide non-trivial improvements over the base algorithm.

Efficient visual event detection using volumetric features

Abstract: This paper studies the use of volumetric features as an alternative to popular local descriptor approaches for event detection in video sequences. Motivated by the recent success of similar ideas in object detection on static images, we generalize the notion of 2D box features to 3D spatio-temporal volumetric features. This general framework enables us to do real-time video analysis. We construct a realtime event detector for each action of interest by learning a cascade of filters based on volumetric features that efficiently scans video sequences in space and time. This event detector recognizes actions that are traditionally problematic for interest point methods - such as smooth motions where insufficient space-time interest points are available. Our experiments demonstrate that the technique accurately detects actions on real-world sequences and is robust to changes in viewpoint, scale and action speed. We also adapt our technique to the related task of human action classification and confirm that it achieves performance comparable to a current interest point based human activity recognizer on a standard database of human activities.

Mapping features to models: a template approach based on superimposed variants

Abstract: Although a feature model can represent commonalities and variabilities in a very concise taxonomic form, features in a feature model are merely symbols. Mapping features to other models, such as behavioral or data specifications, gives them semantics. In this paper, we propose a general template-based approach for mapping feature models to concise representations of variability in different kinds of other models. We show how the approach can be applied to UML 2.0 activity and class models and describe a prototype implementation.

Staged configuration through specialization and multilevel configuration of feature models

Abstract: Feature modeling is a key technique for capturing commonalities and variabilities in system families and product lines. In this article, we propose a cardinality-based notation for feature modeling, which integrates a number of existing extensions of previous approaches. We then introduce and motivate the novel concept of staged configuration. Staged configuration can be achieved by the stepwise specialization of feature models or by multilevel configuration, where the configuration choices available in each stage are defined by separate feature models. Staged configuration is important because, in a realistic development process, different groups and different people make product configuration choices in different stages. Finally, we also discuss how multilevel configuration avoids a breakdown between the different abstraction levels of individual features. This problem, sometimes referred to as 'analysis paralysis', easily occurs in feature modeling because features can denote entities at arbitrary levels of abstraction within a system family. Copyright © 2005 John Wiley & Sons, Ltd.

Ordinal palmprint represention for personal identification [represention read representation]

Abstract: Palmprint-based personal identification, as a new member in the biometrics family, has become an active research topic in recent years. Although great progress has been made, how to represent palmprint for effective classification is still an open problem. In this paper, we present a novel palmprint representation - ordinal measure, which unifies several major existing palmprint algorithms into a general framework. In this framework, a novel palmprint representation method, namely orthogonal line ordinal features, is proposed. The basic idea of this method is to qualitatively compare two elongated, line-like image regions, which are orthogonal in orientation and generate one bit feature code. A palmprint pattern is represented by thousands of ordinal feature codes. In contrast to the state-of-the-art algorithm reported in the literature, our method achieves higher accuracy, with the equal error rate reduced by 42% for a difficult set, while the complexity of feature extraction is halved.

2-D and 3-D Image Registration: for Medical, Remote Sensing, and Industrial Applications

Abstract: Preface. Acknowledgments. Acronyms. 1. Introduction. 2. Preprocessing. 3. Feature Selection. 4. Feature Correspondence. 5. Transformation Functions. 6. Resampling. 7. Performance Evaluation. 8. Image Fusion. 9. Image Mosaicking. 10. Stereo Depth Perception. Glossary. References. Index.

Starburst: A hybrid algorithm for video-based eye tracking combining feature-based and model-based approaches

Abstract: Knowing the user’s point of gaze has significant potential to enhance current human-computer interfaces, given that eye movements can be used as an indicator of the attentional state of a user. The primary obstacle of integrating eye movements into today’s interfaces is the availability of a reliable, low-cost open-source eye-tracking system. Towards making such a system available to interface designers, we have developed a hybrid eye-tracking algorithm that integrates feature-based and model-based approaches and made it available in an open-source package. We refer to this algorithm as "starburst" because of the novel way in which pupil features are detected. This starburst algorithm is more accurate than pure feature-based approaches yet is signi?cantly less time consuming than pure modelbased approaches. The current implementation is tailored to tracking eye movements in infrared video obtained from an inexpensive head-mounted eye-tracking system. A validation study was conducted and showed that the technique can reliably estimate eye position with an accuracy of approximately one degree of visual angle.

Feature-based similarity search in 3D object databases

Abstract: The development of effective content-based multimedia search systems is an important research issue due to the growing amount of digital audio-visual information. In the case of images and video, the growth of digital data has been observed since the introduction of 2D capture devices. A similar development is expected for 3D data as acquisition and dissemination technology of 3D models is constantly improving. 3D objects are becoming an important type of multimedia data with many promising application possibilities. Defining the aspects that constitute the similarity among 3D objects and designing algorithms that implement such similarity definitions is a difficult problem. Over the last few years, a strong interest in methods for 3D similarity search has arisen, and a growing number of competing algorithms for content-based retrieval of 3D objects have been proposed. We survey feature-based methods for 3D retrieval, and we propose a taxonomy for these methods. We also present experimental results, comparing the effectiveness of some of the surveyed methods.

Modeling scenes with local descriptors and latent aspects

Abstract: We present a new approach to model visual scenes in image collections, based on local invariant features and probabilistic latent space models. Our formulation provides answers to three open questions:(l) whether the invariant local features are suitable for scene (rather than object) classification; (2) whether unsupennsed latent space models can be used for feature extraction in the classification task; and (3) whether the latent space formulation can discover visual co-occurrence patterns, motivating novel approaches for image organization and segmentation. Using a 9500-image dataset, our approach is validated on each of these issues. First, we show with extensive experiments on binary and multi-class scene classification tasks, that a bag-of-visterm representation, derived from local invariant descriptors, consistently outperforms state-of-the-art approaches. Second, we show that probabilistic latent semantic analysis (PLSA) generates a compact scene representation, discriminative for accurate classification, and significantly more robust when less training data are available. Third, we have exploited the ability of PLSA to automatically extract visually meaningful aspects, to propose new algorithms for aspect-based image ranking and context-sensitive image segmentation.

Feature level fusion of hand and face biometrics

Abstract: Multibiometric systems utilize the evidence presented by multiple biometric sources (e.g., face and fingerprint, multiple fingers of a user, multiple matchers, etc.) in order to determine or verify the identity of an individual. Information from multiple sources can be consolidated in several distinct levels, including the feature extraction level, match score level and decision level. While fusion at the match score and decision levels have been extensively studied in the literature, fusion at the feature level is a relatively understudied problem. In this paper we discuss fusion at the feature level in 3 different scenarios: (i) fusion of PCA and LDA coefficients of face; (ii) fusion of LDA coefficients corresponding to the R,G,B channels of a face image; (iii) fusion of face and hand modalities. Preliminary results are encouraging and help in highlighting the pros and cons of performing fusion at this level. The primary motivation of this work is to demonstrate the viability of such a fusion and to underscore the importance of pursuing further research in this direction.

CLUE: cluster-based retrieval of images by unsupervised learning

Abstract: In a typical content-based image retrieval (CBIR) system, target images (images in the database) are sorted by feature similarities with respect to the query. Similarities among target images are usually ignored. This paper introduces a new technique, cluster-based retrieval of images by unsupervised learning (CLUE), for improving user interaction with image retrieval systems by fully exploiting the similarity information. CLUE retrieves image clusters by applying a graph-theoretic clustering algorithm to a collection of images in the vicinity of the query. Clustering in CLUE is dynamic. In particular, clusters formed depend on which images are retrieved in response to the query. CLUE can be combined with any real-valued symmetric similarity measure (metric or nonmetric). Thus, it may be embedded in many current CBIR systems, including relevance feedback systems. The performance of an experimental image retrieval system using CLUE is evaluated on a database of around 60,000 images from COREL. Empirical results demonstrate improved performance compared with a CBIR system using the same image similarity measure. In addition, results on images returned by Google's Image Search reveal the potential of applying CLUE to real-world image data and integrating CLUE as a part of the interface for keyword-based image retrieval systems.

An on-line signature verification system based on fusion of local and global information

Abstract: An on-line signature verification system exploiting both local and global information through decision-level fusion is presented. Global information is extracted with a feature-based representation and recognized by using Parzen Windows Classifiers. Local information is extracted as time functions of various dynamic properties and recognized by using Hidden Markov Models. Experimental results are given on the large MCYT signature database (330 signers, 16500 signatures) for random and skilled forgeries. Feature selection experiments based on feature ranking are carried out. It is shown experimentally that the machine expert based on local information outperforms the system based on global analysis when enough training data is available. Conversely, it is found that global analysis is more appropriate in the case of small training set size. The two proposed systems are also shown to give complementary recognition information which is successfully exploited using decision-level score fusion.

fMRI evidence for the neural representation of faces.

Abstract: fMRI (functional magnetic resonance imaging) studies on humans have shown a cortical area, the fusiform face area, that is specialized for face processing. An important question is how faces are represented within this area. This study provides direct evidence for a representation in which individual faces are encoded by their direction (facial identity) and distance (distinctiveness) from a prototypical (mean) face. When facial geometry (head shape, hair line, internal feature size and placement) was varied, the fMRI signal increased with increasing distance from the mean face. Furthermore, adaptation of the fMRI signal showed that the same neural population responds to faces falling along single identity axes within this space.

A sparse object category model for efficient learning and exhaustive recognition

Abstract: We present a "parts and structure" model for object category recognition that can be learnt efficiently and in a semi-supervised manner: the model is learnt from example images containing category instances, without requiring segmentation from background clutter. The model is a sparse representation of the object, and consists of a star topology configuration of parts modeling the output of a variety of feature detectors. The optimal choice of feature types (whose repertoire includes interest points, curves and regions) is made automatically. In recognition, the model may be applied efficiently in an exhaustive manner, bypassing the need for feature detectors, to give the globally optimal match within a query image. The approach is demonstrated on a wide variety of categories, and delivers both successful classification and localization of the object within the image.

Joint Haar-like features for face detection

Abstract: In this paper, we propose a new distinctive feature, called joint Haar-like feature, for detecting faces in images. This is based on co-occurrence of multiple Haar-like features. Feature co-occurrence, which captures the structural similarities within the face class, makes it possible to construct an effective classifier. The joint Haar-like feature can be calculated very fast and has robustness against addition of noise and change in illumination. A face detector is learned by stagewise selection of the joint Haar-like features using AdaBoost. A small number of distinctive features achieve both computational efficiency and accuracy. Experimental results with 5, 676 face images and 30,000 nonface images show that our detector yields higher classification performance than Viola and Jones' detector; which uses a single feature for each weak classifier. Given the same number of features, our method reduces the error by 37%. Our detector is 2.6 times as fast as Viola and Jones' detector to achieve the same performance

Feature-based surface parameterization and texture mapping

Abstract: Surface parameterization is necessary for many graphics tasks: texture-preserving simplification, remeshing, surface painting, and precomputation of solid textures. The stretch caused by a given parameterization determines the sampling rate on the surface. In this article, we present an automatic parameterization method for segmenting a surface into patches that are then flattened with little stretch.Many objects consist of regions of relatively simple shapes, each of which has a natural parameterization. Based on this observation, we describe a three-stage feature-based patch creation method for manifold surfaces. The first two stages, genus reduction and feature identification, are performed with the help of distance-based surface functions. In the last stage, we create one or two patches for each feature region based on a covariance matrix of the feature's surface points.To reduce stretch during patch unfolding, we notice that stretch is a 2 × 2 tensor, which in ideal situations is the identity. Therefore, we use the Green-Lagrange tensor to measure and to guide the optimization process. Furthermore, we allow the boundary vertices of a patch to be optimized by adding scaffold triangles. We demonstrate our feature-based patch creation and patch unfolding methods for several textured models.Finally, to evaluate the quality of a given parameterization, we describe an image-based error measure that takes into account stretch, seams, smoothness, packing efficiency, and surface visibility.

Mining anomalies using traffic feature distributions

Abstract: The increasing practicality of large-scale flow capture makes it possible to conceive of traffic analysis methods that detect and identify a large and diverse set of anomalies. However the challeng...

Feature congestion: a measure of display clutter

Abstract: Management of clutter is an important factor in the design of user interfaces and information visualizations, allowing improved usability and aesthetics. However, clutter is not a well defined concept. In this paper, we present the Feature Congestion measure of display clutter. This measure is based upon extensive modeling of the saliency of elements of a display, and upon a new operational definition of clutter. The current implementation is based upon two features: color and luminance contrast. We have tested this measure on maps that observers ranked by perceived clutter. Results show good agreement between the observers' rankings and our measure of clutter. Furthermore, our measure can be used to make design suggestions in an automated UI critiquing tool.

Integrated circuit layout design methodology with process variation bands

Abstract: A system for analyzing IC layouts and designs by calculating variations of a number of objects to be created on a semiconductor wafer as a result of different process conditions. The variations are analyzed to determine individual feature failures or to rank layout designs by their susceptibility to process variations. In one embodiment, the variations are represented by PV-bands having an inner edge that defines the smallest area in which an object will always print and an outer edge that defines the largest area in which an object will print under some process conditions.

Color2Gray: salience-preserving color removal

Abstract: Visually important image features often disappear when color images are converted to grayscale. The algorithm introduced here reduces such losses by attempting to preserve the salient features of the color image. The Color2Gray algorithm is a 3-step process: 1) convert RGB inputs to a perceptually uniform CIE L*a*b* color space, 2) use chrominance and luminance differences to create grayscale target differences between nearby image pixels, and 3) solve an optimization problem designed to selectively modulate the grayscale representation as a function of the chroma variation of the source image. The Color2Gray results offer viewers salient information missing from previous grayscale image creation methods.