Showing papers on "RANSAC published in 2006"

Visual attention detection in video sequences using spatiotemporal cues

Abstract: Human vision system actively seeks interesting regions in images to reduce the search effort in tasks, such as object detection and recognition. Similarly, prominent actions in video sequences are more likely to attract our first sight than their surrounding neighbors. In this paper, we propose a spatiotemporal video attention detection technique for detecting the attended regions that correspond to both interesting objects and actions in video sequences. Both spatial and temporal saliency maps are constructed and further fused in a dynamic fashion to produce the overall spatiotemporal attention model. In the temporal attention model, motion contrast is computed based on the planar motions (homography) between images, which is estimated by applying RANSAC on point correspondences in the scene. To compensate the non-uniformity of spatial distribution of interest-points, spanning areas of motion segments are incorporated in the motion contrast computation. In the spatial attention model, a fast method for computing pixel-level saliency maps has been developed using color histograms of images. A hierarchical spatial attention representation is established to reveal the interesting points in images as well as the interesting regions. Finally, a dynamic fusion technique is applied to combine both the temporal and spatial saliency maps, where temporal attention is dominant over the spatial model when large motion contrast exists, and vice versa. The proposed spatiotemporal attention framework has been applied on over 20 testing video sequences, and attended regions are detected to highlight interesting objects and motions present in the sequences with very high user satisfaction rate.

Ranking outliers using symmetric neighborhood relationship

Abstract: Mining outliers in database is to find exceptional objects that deviate from the rest of the data set. Besides classical outlier analysis algorithms, recent studies have focused on mining local outliers, i.e., the outliers that have density distribution significantly different from their neighborhood. The estimation of density distribution at the location of an object has so far been based on the density distribution of its k-nearest neighbors [2,11]. However, when outliers are in the location where the density distributions in the neighborhood are significantly different, for example, in the case of objects from a sparse cluster close to a denser cluster, this may result in wrong estimation. To avoid this problem, here we propose a simple but effective measure on local outliers based on a symmetric neighborhood relationship. The proposed measure considers both neighbors and reverse neighbors of an object when estimating its density distribution. As a result, outliers so discovered are more meaningful. To compute such local outliers efficiently, several mining algorithms are developed that detects top-n outliers based on our definition. A comprehensive performance evaluation and analysis shows that our methods are not only efficient in the computation but also more effective in ranking outliers.

Ranking outliers using symmetric neighborhood relationship

Abstract: Mining outliers in database is to find exceptional objects that deviate from the rest of the data set. Besides classical outlier analysis algorithms, recent studies have focused on mining local outliers, i.e., the outliers that have density distribution significantly different from their neighborhood. The estimation of density distribution at the location of an object has so far been based on the density distribution of its k-nearest neighbors [2,11]. However, when outliers are in the location where the density distributions in the neighborhood are significantly different, for example, in the case of objects from a sparse cluster close to a denser cluster, this may result in wrong estimation. To avoid this problem, here we propose a simple but effective measure on local outliers based on a symmetric neighborhood relationship. The proposed measure considers both neighbors and reverse neighbors of an object when estimating its density distribution. As a result, outliers so discovered are more meaningful. To compute such local outliers efficiently, several mining algorithms are developed that detects top-n outliers based on our definition. A comprehensive performance evaluation and analysis shows that our methods are not only efficient in the computation but also more effective in ranking outliers.

Distance-based detection and prediction of outliers

Abstract: A distance-based outlier detection method that finds the top outliers in an unlabeled data set and provides a subset of it, called outlier detection solving set, that can be used to predict the outlierness of new unseen objects, is proposed. The solving set includes a sufficient number of points that permits the detection of the top outliers by considering only a subset of all the pairwise distances from the data set. The properties of the solving set are investigated, and algorithms for computing it, with subquadratic time requirements, are proposed. Experiments on synthetic and real data sets to evaluate the effectiveness of the approach are presented. A scaling analysis of the solving set size is performed, and the false positive rate, that is, the fraction of new objects misclassified as outliers using the solving set instead of the overall data set, is shown to be negligible. Finally, to investigate the accuracy in separating outliers from inliers, ROC analysis of the method is accomplished. Results obtained show that using the solving set instead of the data set guarantees a comparable quality of the prediction, but at a lower computational cost.

Structure from motion with wide circular field of view cameras

Abstract: This paper presents a method for fully automatic and robust estimation of two-view geometry, autocalibration, and 3D metric reconstruction from point correspondences in images taken by cameras with wide circular field of view. We focus on cameras which have more than 180deg field of view and for which the standard perspective camera model is not sufficient, e.g., the cameras equipped with circular fish-eye lenses Nikon FC-E8 (183deg), Sigma 8 mm-f4-EX (180deg), or with curved conical mirrors. We assume a circular field of view and axially symmetric image projection to autocalibrate the cameras. Many wide field of view cameras can still be modeled by the central projection followed by a nonlinear image mapping. Examples are the above-mentioned fish-eye lenses and properly assembled catadioptric cameras with conical mirrors. We show that epipolar geometry of these cameras can be estimated from a small number of correspondences by solving a polynomial eigenvalue problem. This allows the use of efficient RANSAC robust estimation to find the image projection model, the epipolar geometry, and the selection of true point correspondences from tentative correspondences contaminated by mismatches. Real catadioptric cameras are often slightly noncentral. We show that the proposed autocalibration with approximate central models is usually good enough to get correct point correspondences which can be used with accurate noncentral models in a bundle adjustment to obtain accurate 3D scene reconstruction. Noncentral camera models are dealt with and results are shown for catadioptric cameras with parabolic and spherical mirrors

RANSAC for (Quasi-)Degenerate data (QDEGSAC)

Abstract: The computation of relations from a number of potential matches is a major task in computer vision. Often RANSAC is employed for the robust computation of relations such as the fundamental matrix. For (quasi-)degenerate data however, it often fails to compute the correct relation. The computed relation is always consistent with the data but RANSAC does not verify that it is unique. The paper proposes a framework that estimates the correct relation with the same robustness as RANSAC even for (quasi-)degenerate data. The approach is based on a hierarchical RANSAC over the number of constraints provided by the data. In contrast to all previously presented algorithms for (quasi-)degenerate data our technique does not require problem specific tests or models to deal with degenerate configurations. Accordingly it can be applied for the estimation of any relation on any data and is not limited to a special type of relation as previous approaches. The results are equivalent to the results achieved by state of the art approaches that employ knowledge about degeneracies.

Detection and classification of highway lanes using vehicle motion trajectories

Abstract: Intelligent vision-based traffic surveillance systems are assuming an increasingly important role in highway monitoring and road management schemes. This paper describes a low-level object tracking system that produces accurate vehicle motion trajectories that can be further analyzed to detect lane centers and classify lane types. Accompanying techniques for indexing and retrieval of anomalous trajectories are also derived. The predictive trajectory merge-and-split algorithm is used to detect partial or complete occlusions during object motion and incorporates a Kalman filter that is used to perform vehicle tracking. The resulting motion trajectories are modeled using variable low-degree polynomials. A K-means clustering technique on the coefficient space can be used to obtain approximate lane centers. Estimation bias due to vehicle lane changes can be removed using robust estimation techniques based on Random Sample Consensus (RANSAC). Through the use of nonmetric distance functions and a simple directional indicator, highway lanes can be classified into one of the following categories: entry, exit, primary, or secondary. Experimental results are presented to show the real-time application of this approach to multiple views obtained by an uncalibrated pan-tilt-zoom traffic camera monitoring the junction of two busy intersecting highways.

Detecting outliers using transduction and statistical testing

Abstract: Outlier detection can uncover malicious behavior in fields like intrusion detection and fraud analysis. Although there has been a significant amount of work in outlier detection, most of the algorithms proposed in the literature are based on a particular definition of outliers (e.g., density-based), and use ad-hoc thresholds to detect them. In this paper we present a novel technique to detect outliers with respect to an existing clustering model. However, the test can also be successfully utilized to recognize outliers when the clustering information is not available. Our method is based on Transductive Confidence Machines, which have been previously proposed as a mechanism to provide individual confidence measures on classification decisions. The test uses hypothesis testing to prove or disprove whether a point is fit to be in each of the clusters of the model. We experimentally demonstrate that the test is highly robust, and produces very few misdiagnosed points, even when no clustering information is available. Furthermore, our experiments demonstrate the robustness of our method under the circumstances of data contaminated by outliers. We finally show that our technique can be successfully applied to identify outliers in a noisy data set for which no information is available (e.g., ground truth, clustering structure, etc.). As such our proposed methodology is capable of bootstrapping from a noisy data set a clean one that can be used to identify future outliers.

Multiple Sound Source Mapping for a Mobile Robot by Self-motion Triangulation

Abstract: The paper describes a 2D sound source mapping system for a mobile robot. We developed a multiple sound sources localization method for a mobile robot with a 32 channel concentric microphone array. The system can separate multiple moving sound sources using direction localization. Directional localization and separation of different pressure sound sources is achieved using the delay and sum beam forming (DSBF) and the frequency band selection (FBS) algorithm. Sound sources were mapped by using a wheeled robot equipped with the microphone array. The robot localizes sounds direction on the move and estimates sound sources position using triangulation. Assuming the movement of sound sources, the system set a time limit and uses only the last few seconds data. By using the random sample consensus (RANSAC) algorithm for position estimation, we achieved 2D multiple sound source mapping from time limited data with high accuracy. Also, moving sound source separation is experimentally demonstrated with segments of the DSBF enhanced signal derived from the localization process

Beyond RANSAC: User Independent Robust Regression

Abstract: RANSAC is the most widely used robust regression algorithm in computer vision. However, RANSAC has a few drawbacks which make it difficult to use in a lot of applications. Some of these problems have been addressed through improved sampling algorithms or better cost functions, but an important problem still remains. The algorithms are not user independent, and require some knowledge of the scale of the inlier noise. The projection based M-estimator (pbM) offers a solution to this by reframing the regression problem in a projection pursuit framework. In this paper we derive the pbM algorithm for heteroscedastic data. Our algorithm is applied to various real problems and its performance is compared with RANSAC and MSAC. It is shown that pbM gives better results than RANSAC and MSAC in spite of being user independent.

Realtime lane tracking of curved local road

Abstract: A lane detection system is an important component of many intelligent transportation systems. We present a robust realtime lane tracking algorithm for a curved local road. First, we present a comparative study to find a good realtime lane marking classifier. Once lane markings are detected, they are grouped into many lane boundary hypotheses represented by constrained cubic spline curves. We present a robust hypothesis generation algorithm using a particle filtering technique and a RANSAC (random sample concensus) algorithm. We introduce a probabilistic approach to group lane boundary hypotheses into left and right lane boundaries. The proposed grouping approach can be applied to general part-based object tracking problems. It incorporates a likelihood-based object recognition technique into a Markov-style process. An experimental result on local streets shows that the suggested algorithm is very reliable

Automatic relative orientation of images

Abstract: This paper presents a new approach to full automatic relative orientation of several digital images taken with a calibrated camera. This approach uses new algorithms for feature extraction and relative orientation developed in the last few years. There is no need for special markers in the scene nor for approximate values for the parameters of the exterior orientation. We use the point operator developed by D. G. Lowe (Lowe, 2004), which extracts points with scaleand rotation-invariant descriptors (SIFT-features). These descriptors allow a successful matching of image points even in situations with highly convergent images. The approach consists of the following steps: After extracting image points on all images each image pair is matched using the SIFT parameters only. No prior information about the pose of the images or the overlapping parts of the images is needed. For every image pair a relative orientation is computed using a RANSAC procedure. Here we use the new 5-point algorithm developed by D. Nister (Nister, 2004). Based on these orientations approximate values for the orientation parameters and the object coordinates are calculated. This is achieved by computing the relative scale and transforming into a common coordinate system. Several tests are carried out to ensure reliable inputs for the currently final step: a bundle block adjustment. The paper discusses the practical impacts of the algorithms involved. Examples of different indoorand outdoor-scenes including a dataset of tilted aerial images are presented and the results of the approach are evaluated. These results show that the approach can be used for a wide range of scenes with different types of the image geometry and taken with different types of cameras including inexpensive consumer cameras. In particular we investigate in the robustness of the algorithms, e.g. in geometric tests on image triplets. In the outlook further developments like the use of image pyramids with a modified matching are discussed. 1 Published in Proceedings of the 5th Turkish-German Joint Geodetic Days, March 29 – 31, 2006, Berlin, ISBN 3-9809030-4-4

Epipolar Geometry from Two Correspondences

Abstract: A novel algorithm for robust RANSAC-like estimation of epipolar geometry (of uncalibrated camera pair) from two correspondences of local affine frames (LAFs) is presented. Each LAF is constructed from three points independently detected on a maximally stable extremal region. The algorithm assumes that a sufficiently accurate approximation of the fundamental matrix is obtained from two LAF correspondences by the 6-point algorithm of Stewe´nius et al. The so-far-the-best hypotheses are further processed by so-called local optimization to estimate the epipolar geometry. Special attention is paid to planar sample degeneracy, since the probability of drawing two coplanar LAF correspondences is not negligible. Combining the 6-point solver, local optimization, and the degeneracy test enables RANSAC to draw samples of only two LAFs to generate hypotheses and thus to reduce the number of samples drawn. We experimentally show that using the 6-point algorithm (approximating the real camera by camera with unit aspect ratio, zero skew, principal point in the center of image, and a common unknown focal length) generates hypotheses that are sufficient for EG estimation in LO-RANSAC framework.

Real-time upper body detection and 3d pose estimation in monoscopic images

Abstract: This paper presents a novel solution to the difficult task of both detecting and estimating the 3D pose of humans in monoscopic images. The approach consists of two parts. Firstly the location of a human is identified by a probabalistic assembly of detected body parts. Detectors for the face, torso and hands are learnt using adaBoost. A pose likliehood is then obtained using an a priori mixture model on body configuration and possible configurations assembled from available evidence using RANSAC. Once a human has been detected, the location is used to initialise a matching algorithm which matches the silhouette and edge map of a subject with a 3D model. This is done efficiently using chamfer matching, integral images and pose estimation from the initial detection stage. We demonstrate the application of the approach to large, cluttered natural images and at near framerate operation (16fps) on lower resolution video streams.

Generalized RANSAC Framework for Relaxed Correspondence Problems

Abstract: Finding correspondences between two (widely) separated views is essential for several computer vision tasks, such as structure and motion estimation and object recognition. In the wide-baseline matching using scale and/or affine invariant features the search for correspondences typically proceeds in two stages. In the first stage a putative set of correspondences is obtained based on distances between feature descriptors. In the second stage the matches are refined by imposing global geometric constraints by means of robust estimation of the epipolar geometry and the incorrect matches are rejected as outliers. For a feature in one view, usually only one "best" feature (the nearest neighbor) in the other view is chosen as corresponding feature, despite the fact that several match candidates exist. In this paper, we will consider multiple candidate matches for each feature, and integrate this choice with the robust estimation stage, thus avoiding the early commitment to the "best" one. This yields a generalized RANSAC framework for identifying the true correspondences among sets of matches. We examine the effectiveness of different sampling strategies for sets of correspondences and test the approach extensively using real examples of hard correspondence problems caused by a large motion between views and/or ambiguities due to repetitive scene structures.

RANBAR: RANSAC-based resilient aggregation in sensor networks

Abstract: We present a novel outlier elimination technique designed for sensor networks. This technique is called RANBAR and it is based on the RANSAC (RANdom SAmple Consensus) paradigm, which is well-known in computer vision and in automated cartography. The RANSAC paradigm gives us a hint on how to instantiate a model if there are a lot of compromised data elements.However,the paradigm does not specify an algorithm and it uses a guess for the number of compromised elements, which is not known in general in real life environments. We developed the RANBAR algorithm following this paradigm and we eliminated the need for the guess. Our RANBAR algorithm is therefore capable to handle a high percent of outlier measurement data by leaning on only one preassumption,namely that the sample is i.i.d. in the unattacked case. We implemented the algorithm in a simulation environment and we used it to filter out outlier elements from a sample before an aggregation procedure. The aggregation function that we used was the average. We show that the algorithm guarantees a small distortion on the output of the aggregator even if almost half of the sample is compromised. Compared to other resilient aggregation algorithms, like the trimmed average and the median, our RANBAR algorithm results in smaller distortion, especially for high attack strengths.

Data Association Using Visual Object Recognition for EKF-SLAM in Home Environment

Abstract: Reliable data association is crucial to localization and map building for mobile robot applications. For that reason, many mobile robots tend to choose vision-based SLAM solutions. In this paper, a SLAM scheme based on visual object recognition, not just a scene matching, in home environment is proposed without using artificial landmarks. For the object-based SLAM, the following algorithms are suggested: 1) a novel local invariant feature extraction by combining advantages of multi-scale Harris corner as a detector and its SIFT descriptor for natural object recognition, 2) the RANSAC clustering for robust object recognition in the presence of outliers and 3) calculating accurate metric information for SLAM update. The proposed algorithms increase robustness by correct data association and accurate observation. Moreover, it also can be easily implemented real-time by reducing the number of representative landmarks, i.e. objects. The performance of the proposed algorithm was verified by experiments using EKF-SLAM with a stereo camera in home-like environments, and it showed that the final pose error was bounded after battery-run-out autonomous navigation for 50 minutes

Experiments with robust estimation techniques in real-time robot vision

Abstract: The goal of this paper is to present an overview of robust estimation techniques with a special focus on robotic vision applications. In this particular context, constraints due computation time have to be considered in the choice of the estimation algorithm. Among the numerous techniques proposed in the literature to obtained robust estimation we have, not being exhaustive, Hough transform, RANSAC (Random Sample Consensus), the LMedS (Least Median of Squares), the Mestimators, estimators, etc. In this overview, we describe these various approaches in the light of a simple example. Finally, we illustrate the use of robust estimation techniques by various examples in real-time robot vision.

A Feature-based Approach for Dense Segmentation and Estimation of Large Disparity Motion

Abstract: We present a novel framework for motion segmentation that combines the concepts of layer-based methods and feature-based motion estimation. We estimate the initial correspondences by comparing vectors of filter outputs at interest points, from which we compute candidate scene relations via random sampling of minimal subsets of correspondences. We achieve a dense, piecewise smooth assignment of pixels to motion layers using a fast approximate graphcut algorithm based on a Markov random field formulation. We demonstrate our approach on image pairs containing large inter-frame motion and partial occlusion. The approach is efficient and it successfully segments scenes with inter-frame disparities previously beyond the scope of layer-based motion segmentation methods. We also present an extension that accounts for the case of non-planar motion, in which we use our planar motion segmentation results as an initialization for a regularized Thin Plate Spline fit. In addition, we present applications of our method to automatic object removal and to structure from motion.

Real-time lane detection in various conditions and night cases

Abstract: In this paper, we propose a real-time lane detection algorithm based on a hyperbola-pair lane boundary model and an improved RANSAC paradigm. Instead of modeling each road boundary separately, we propose a model to describe the road boundary as a pair of parallel hyperbolas on the ground plane. A fuzzy measurement is introduced into the RANSAC paradigm to improve the accuracy and robustness of fitting the points on the boundaries into the model. Our method is able to deal with existence of partial occlusion, other traffic participants and markings, etc. Experiment in many different conditions, including various conditions of illumination, weather and road, demonstrates its high performance and accuracy

Outlier Detection and Editing Procedures for Continuous Multivariate Data

Abstract: We present a semi-automatic method of outlier detection for continuous, multivariate survey data that is designed to identify outlying cases and suggest potential errors on a case-by-case basis, in the presence of missing data.

Matching slides to presentation videos using SIFT and scene background matching

Abstract: We present a general approach for automatically matching electronic slides to videos of corresponding presentations for use in distance learning and video proceedings of conferences. We deal with a large variety of videos, various frame compositions and color balances, arbitrary slides sequence and with dynamic cameras switching, pan, tilt and zoom. To achieve high accuracy, we develop a two-phases process with unsupervised scene background modelling. In the first phase, scale invariant feature transform (SIFT) keypoints are applied to frame to slide matching, under constraint projective transformation (constraint homography) using a random sample consensus (RANSAC). Successful first-phase matches are then used to automatically build a scene background model. In the second phase the background model is applied to the remaining unmatched frames to boost the matching performance for difficult cases such as wide field of view camera shots where the slide shows as a small portion of the frame. We also show that color correction is helpful when color-related similarity measures are used for identifying slides. We provide detailed quantitative experimentation results characterizing the effect of each part of our approach. The results show that our approach is robust and achieves high performance on matching slides to a number of videos with different styles.

Stereo Tracking and Three-Point/One-Point Algorithms - A Robust Approach in Visual Odometry

Abstract: In this paper, we present an approach of calculating visual odometry for outdoor robots equipped with a stereo rig. Instead of the typical feature matching or tracking, we use an improved stereo-tracking method that simultaneously decides the feature displacement in both cameras. Based on the matched features, a three-point algorithm for the resulting quadrifocal setting is carried out in a RANSAC framework to recover the unknown odometry. In addition, the change in rotation can be derived from infinity homography, and the remaining translational unknowns can be obtained even faster consequently . Both approaches are quite robust and deal well with challenging conditions such as wheel slippage.

Estimating the essential matrix: goodsac versus ransac

Abstract: GOODSAC is a paradigm for estimation of model parameters given measurements that are contaminated by outliers. Thus, it is an alternative to the well known RANSAC strategy. GOODSAC’s search for a proper set of inliers does not only maximize the sheer size of this set, but also takes other assessments for the utility into account. Assessments can be used on many levels of the process to control the search and foster precision and proper utilization of the computational resources. This contribution discusses and compares the two methods. In particular, the estimation of essential matrices is used as example. The comparison is performed on synthetic and real data and is based on standard statistical methods, where GOODSAC achieves higher precision than RANSAC.

Learning efficient linear predictors for motion estimation

Abstract: A novel object representation for tracking is proposed. The tracked object is represented as a constellation of spatially localised linear predictors which are learned on a single training image. In the learning stage, sets of pixels whose intensities allow for optimal least square predictions of the transformations are selected as a support of the linear predictor. The approach comprises three contributions: learning object specific linear predictors, explicitly dealing with the predictor precision computational complexity trade-off and selecting a view-specific set of predictors suitable for global object motion estimate. Robustness to occlusion is achieved by RANSAC procedure. The learned tracker is very efficient, achieving frame rate generally higher than 30 frames per second despite the Matlab implementation.

Incremental RANSAC for online relocation in large dynamic environments

Abstract: Vehicle relocation is the problem in which a mobile robot has to estimate the self-position with respect to an a priori map of landmarks using the perception and the motion measurements without using any knowledge of the initial self-position. Recently, random sample consensus (RANSAC), a robust multi-hypothesis estimator, has been successfully applied to offline relocation in static environments. On the other hand, online relocation in dynamic environments is still a difficult problem, for available computation time is always limited, and for measurement include many outliers. To realize real time algorithm for such an online process, we have developed an incremental version of RANSAC algorithm by extending an efficient preemption RANSAC scheme. This novel scheme named incremental RANSAC is able to find inlier hypotheses of self-positions out of large number of outlier hypotheses contaminated by outlier measurements

Articulated motion segmentation using RANSAC with priors

Abstract: Articulated motions are partially dependent. Most of the existing segmentation methods, e.g. Costeira and Kanade[2], can not be applied to articulated motions. We propose a novel algorithm for articulated motion segmentation called RANSAC with priors. It does not require prior knowledge of the number of articulated parts. It is both robust and efficient. Its robustness comes from its RANSAC nature. Its efficiency is due to the priors, which are derived from the spectral affinities between every pair of trajectories. We test our algorithm with synthetic and real data. In some highly challenging case, where other motion segmentation algorithms may fail, our algorithm still achieves robust results. Though our algorithm is inspired by articulated motions, it also applies to independent motions which can be regarded as a special case and treated uniformly.

Robust structure and motion from outlines of smooth curved surfaces

Abstract: This paper addresses the problem of estimating the motion of a camera as it observes the outline (or apparent contour) of a solid bounded by a smooth surface in successive image frames. In this context, the surface points that project onto the outline of an object depend on the viewpoint and the only true correspondences between two outlines of the same object are the projections of frontier points where the viewing rays intersect in the tangent plane of the surface. In turn, the epipolar geometry is easily estimated once these correspondences have been identified. Given the apparent contours detected in an image sequence, a robust procedure based on RANSAC and a voting strategy is proposed to simultaneously estimate the camera configurations and a consistent set of frontier point projections by enforcing the redundancy of multiview epipolar geometry. The proposed approach is, in principle, applicable to orthographic, weak-perspective, and affine projection models. Experiments with nine real image sequences are presented for the orthographic projection case, including a quantitative comparison with the ground-truth data for the six data sets for which the latter information is available. Sample visual hulls have been computed from all image sequences for qualitative evaluation.

Learning efficient linear predictors for motion estimation

Abstract: A novel object representation for tracking is proposed. The tracked object is represented as a constellation of spatially localised linear predictors which are learned on a single training image. In the learning stage, sets of pixels whose intensities allow for optimal least square predictions of the transformations are selected as a support of the linear predictor. The approach comprises three contributions: learning object specific linear predictors, explicitly dealing with the predictor precision – computational complexity trade-off and selecting a view-specific set of predictors suitable for global object motion estimate. Robustness to occlusion is achieved by RANSAC procedure. The learned tracker is very efficient, achieving frame rate generally higher than 30 frames per second despite the Matlab implementation.

Subspace Estimation Using Projection Based M-Estimators over Grassmann Manifolds

Abstract: We propose a solution to the problem of robust subspace estimation using the projection based M-estimator. The new method handles more outliers than inliers, does not require a user defined scale of the noise affecting the inliers, handles noncentered data and nonorthogonal subspaces. Other robust methods like RANSAC, use an input for the scale, while methods for subspace segmentation, like GPCA, are not robust. Synthetic data and three real cases of multibody factorization show the superiority of our method, in spite of user independence.