Pub Date : 2008-06-23DOI: 10.1109/CVPRW.2008.4562959
R. Raguram, S. Lazebnik
This paper considers the problem of selecting iconic images to summarize general visual categories. We define iconic images as high-quality representatives of a large group of images consistent both in appearance and semantics. To find such groups, we perform joint clustering in the space of global image descriptors and latent topic vectors of tags associated with the images. To select the representative iconic images for the joint clusters, we use a quality ranking learned from a large collection of labeled images. For the purposes of visualization, iconic images are grouped by semantic ldquothemerdquo and multidimensional scaling is used to compute a 2D layout that reflects the relationships between the themes. Results on four large-scale datasets demonstrate the ability of our approach to discover plausible themes and recurring visual motifs for challenging abstract concepts such as ldquoloverdquo and ldquobeautyrdquo.
{"title":"Computing iconic summaries of general visual concepts","authors":"R. Raguram, S. Lazebnik","doi":"10.1109/CVPRW.2008.4562959","DOIUrl":"https://doi.org/10.1109/CVPRW.2008.4562959","url":null,"abstract":"This paper considers the problem of selecting iconic images to summarize general visual categories. We define iconic images as high-quality representatives of a large group of images consistent both in appearance and semantics. To find such groups, we perform joint clustering in the space of global image descriptors and latent topic vectors of tags associated with the images. To select the representative iconic images for the joint clusters, we use a quality ranking learned from a large collection of labeled images. For the purposes of visualization, iconic images are grouped by semantic ldquothemerdquo and multidimensional scaling is used to compute a 2D layout that reflects the relationships between the themes. Results on four large-scale datasets demonstrate the ability of our approach to discover plausible themes and recurring visual motifs for challenging abstract concepts such as ldquoloverdquo and ldquobeautyrdquo.","PeriodicalId":102206,"journal":{"name":"2008 IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-06-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130805529","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2008-06-23DOI: 10.1109/CVPRW.2008.4563139
C. Beder, Richard Steffen
We will present a novel incremental algorithm for the task of online least-squares estimation. Our approach aims at combining the accuracy of least-squares estimation and the fast computation of recursive estimation techniques like the Kalman filter. Analyzing the structure of least-squares estimation we devise a novel incremental algorithm, which is able to introduce new unknown parameters and observations into an estimation simultaneously and is equivalent to the optimal overall estimation in case of linear models. It constitutes a direct generalization of the well-known Kalman filter allowing to augment the state vector inside the update step. In contrast to classical recursive estimation techniques no artificial initial covariance for the new unknown parameters is required here. We will show, how this new algorithm allows more flexible parameter estimation schemes especially in the case of scene and motion reconstruction from image sequences. Since optimality is not guaranteed in the non-linear case we will also compare our incremental estimation scheme to the optimal bundle adjustment on a real image sequence. It will be shown that competitive results are achievable using the proposed technique.
{"title":"Incremental estimation without specifying a-priori covariance matrices for the novel parameters","authors":"C. Beder, Richard Steffen","doi":"10.1109/CVPRW.2008.4563139","DOIUrl":"https://doi.org/10.1109/CVPRW.2008.4563139","url":null,"abstract":"We will present a novel incremental algorithm for the task of online least-squares estimation. Our approach aims at combining the accuracy of least-squares estimation and the fast computation of recursive estimation techniques like the Kalman filter. Analyzing the structure of least-squares estimation we devise a novel incremental algorithm, which is able to introduce new unknown parameters and observations into an estimation simultaneously and is equivalent to the optimal overall estimation in case of linear models. It constitutes a direct generalization of the well-known Kalman filter allowing to augment the state vector inside the update step. In contrast to classical recursive estimation techniques no artificial initial covariance for the new unknown parameters is required here. We will show, how this new algorithm allows more flexible parameter estimation schemes especially in the case of scene and motion reconstruction from image sequences. Since optimality is not guaranteed in the non-linear case we will also compare our incremental estimation scheme to the optimal bundle adjustment on a real image sequence. It will be shown that competitive results are achievable using the proposed technique.","PeriodicalId":102206,"journal":{"name":"2008 IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops","volume":"73 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-06-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124276834","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2008-06-23DOI: 10.1109/CVPRW.2008.4563107
Yan Xiang, G. Su
Information fusion of multi-biometrics has become a center of focus for biometrics based identification and verification, and there are two fusion categories: intra-modal fusion and multi-modal fusion. In this paper, an intra-modal fusion, that is, multi-parts and multi-feature fusion (MPMFF) for face verification is studied. Two face representations are exploited, including the gray-level intensity feature and Gabor feature. Different from most face recognition methods, the MPMFF method divides a face image into five parts: bare face, eyebrows, eyes, nose and mouth, and different features of the same face part are fused at feature level. Then at decision level, five matching results based on the combined-features of different parts are calculated into a final similar score according to the weighted sum rule. Experiment results on FERET face database and our own face database show that the multi-parts and multi-feature fusion method improves the face verification performance.
{"title":"Multi-parts and multi-feature fusion in face verification","authors":"Yan Xiang, G. Su","doi":"10.1109/CVPRW.2008.4563107","DOIUrl":"https://doi.org/10.1109/CVPRW.2008.4563107","url":null,"abstract":"Information fusion of multi-biometrics has become a center of focus for biometrics based identification and verification, and there are two fusion categories: intra-modal fusion and multi-modal fusion. In this paper, an intra-modal fusion, that is, multi-parts and multi-feature fusion (MPMFF) for face verification is studied. Two face representations are exploited, including the gray-level intensity feature and Gabor feature. Different from most face recognition methods, the MPMFF method divides a face image into five parts: bare face, eyebrows, eyes, nose and mouth, and different features of the same face part are fused at feature level. Then at decision level, five matching results based on the combined-features of different parts are calculated into a final similar score according to the weighted sum rule. Experiment results on FERET face database and our own face database show that the multi-parts and multi-feature fusion method improves the face verification performance.","PeriodicalId":102206,"journal":{"name":"2008 IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops","volume":"72 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-06-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124447424","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2008-06-23DOI: 10.1109/CVPRW.2008.4562960
Grant Schindler, Larry Zitnick, Matthew A. Brown
In this paper, we examine the problem of internet video categorization. Specifically, we explore the representation of a video as a ldquobag of wordsrdquo using various combinations of spatial and temporal descriptors. The descriptors incorporate both spatial and temporal gradients as well as optical flow information. We achieve state-of-the-art results on a standard human activity recognition database and demonstrate promising category recognition performance on two new databases of approximately 1000 and 1500 online user-submitted videos, which we will be making available to the community.
{"title":"Internet video category recognition","authors":"Grant Schindler, Larry Zitnick, Matthew A. Brown","doi":"10.1109/CVPRW.2008.4562960","DOIUrl":"https://doi.org/10.1109/CVPRW.2008.4562960","url":null,"abstract":"In this paper, we examine the problem of internet video categorization. Specifically, we explore the representation of a video as a ldquobag of wordsrdquo using various combinations of spatial and temporal descriptors. The descriptors incorporate both spatial and temporal gradients as well as optical flow information. We achieve state-of-the-art results on a standard human activity recognition database and demonstrate promising category recognition performance on two new databases of approximately 1000 and 1500 online user-submitted videos, which we will be making available to the community.","PeriodicalId":102206,"journal":{"name":"2008 IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops","volume":"4 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-06-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116952599","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2008-06-23DOI: 10.1109/CVPRW.2008.4562982
Micha Feigin-Almon, N. Sochen, B. Vemuri
The Laplacian raised to fractional powers can be used to generate scale spaces as was shown in recent literature. This was later extended for inhomogeneous diffusion processes and more general functions of the Laplacian and studied for the Perona-Malik case. In this paper we extend the results to the truly anisotropic Beltrami flow. We additionally introduce a technique for splitting up the work into smaller patches of the image which greatly reduce the computational complexity and allow for the parallelization of the algorithm. Important issues involved in the numerical implementation are discussed.
{"title":"Efficient anisotropic α-Kernels decompositions and flows","authors":"Micha Feigin-Almon, N. Sochen, B. Vemuri","doi":"10.1109/CVPRW.2008.4562982","DOIUrl":"https://doi.org/10.1109/CVPRW.2008.4562982","url":null,"abstract":"The Laplacian raised to fractional powers can be used to generate scale spaces as was shown in recent literature. This was later extended for inhomogeneous diffusion processes and more general functions of the Laplacian and studied for the Perona-Malik case. In this paper we extend the results to the truly anisotropic Beltrami flow. We additionally introduce a technique for splitting up the work into smaller patches of the image which greatly reduce the computational complexity and allow for the parallelization of the algorithm. Important issues involved in the numerical implementation are discussed.","PeriodicalId":102206,"journal":{"name":"2008 IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops","volume":"22 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-06-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123611112","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2008-06-23DOI: 10.1109/CVPRW.2008.4563136
Vidya N. Murali, Stan Birchfield
An algorithm is proposed to answer the challenges of autonomous corridor navigation and mapping by a mobile robot equipped with a single forward-facing camera. Using a combination of corridor ceiling lights, visual homing, and entropy, the robot is able to perform straight line navigation down the center of an unknown corridor. Turning at the end of a corridor is accomplished using Jeffrey divergence and time-to-collision, while deflection from dead ends and blank walls uses a scalar entropy measure of the entire image. When combined, these metrics allow the robot to navigate in both textured and untextured environments. The robot can autonomously explore an unknown indoor environment, recovering from difficult situations like corners, blank walls, and initial heading toward a wall. While exploring, the algorithm constructs a Voronoi-based topo-geometric map with nodes representing distinctive places like doors, water fountains, and other corridors. Because the algorithm is based entirely upon low-resolution (32 times 24) grayscale images, processing occurs at over 1000 frames per second.
{"title":"Autonomous navigation and mapping using monocular low-resolution grayscale vision","authors":"Vidya N. Murali, Stan Birchfield","doi":"10.1109/CVPRW.2008.4563136","DOIUrl":"https://doi.org/10.1109/CVPRW.2008.4563136","url":null,"abstract":"An algorithm is proposed to answer the challenges of autonomous corridor navigation and mapping by a mobile robot equipped with a single forward-facing camera. Using a combination of corridor ceiling lights, visual homing, and entropy, the robot is able to perform straight line navigation down the center of an unknown corridor. Turning at the end of a corridor is accomplished using Jeffrey divergence and time-to-collision, while deflection from dead ends and blank walls uses a scalar entropy measure of the entire image. When combined, these metrics allow the robot to navigate in both textured and untextured environments. The robot can autonomously explore an unknown indoor environment, recovering from difficult situations like corners, blank walls, and initial heading toward a wall. While exploring, the algorithm constructs a Voronoi-based topo-geometric map with nodes representing distinctive places like doors, water fountains, and other corridors. Because the algorithm is based entirely upon low-resolution (32 times 24) grayscale images, processing occurs at over 1000 frames per second.","PeriodicalId":102206,"journal":{"name":"2008 IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops","volume":" 17","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-06-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"120834117","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2008-06-23DOI: 10.1109/CVPRW.2008.4563007
Yun Zhu, X. Papademetris, A. Sinusas, J. Duncan
In this paper we propose an integrated cardiac segmentation and motion tracking algorithm. First, we present a subject-specific dynamical model (SSDM) that simultaneously handles inter-subject variability and temporal dynamics (intra-subject variability), such that it can progressively identify the subject vector associated with a new cardiac sequence, and use this subject vector to predict the subject-specific segmentation of the future frames based on the shapes observed in earlier frames. Second, we use the segmentation as a guide in selecting feature points with significant shape characteristics, and invoke the generalized robust point matching (G-RPM) strategy with boundary element method (BEM)-based regularization model to estimate physically realistic displacement field in a computationally efficient way. The integrated algorithm is formulated in a recursive Bayesian framework that sequentially segments cardiac images and estimates myocardial displacements. ldquoLeave-one-outrdquo validation on 32 sequences demonstrates that the segmentation results are improved when the SSDM is used, and the tracking results are much more accurate when the segmentation module is added.
{"title":"Integrated segmentation and motion analysis of cardiac MR images using a subject-specific dynamical model","authors":"Yun Zhu, X. Papademetris, A. Sinusas, J. Duncan","doi":"10.1109/CVPRW.2008.4563007","DOIUrl":"https://doi.org/10.1109/CVPRW.2008.4563007","url":null,"abstract":"In this paper we propose an integrated cardiac segmentation and motion tracking algorithm. First, we present a subject-specific dynamical model (SSDM) that simultaneously handles inter-subject variability and temporal dynamics (intra-subject variability), such that it can progressively identify the subject vector associated with a new cardiac sequence, and use this subject vector to predict the subject-specific segmentation of the future frames based on the shapes observed in earlier frames. Second, we use the segmentation as a guide in selecting feature points with significant shape characteristics, and invoke the generalized robust point matching (G-RPM) strategy with boundary element method (BEM)-based regularization model to estimate physically realistic displacement field in a computationally efficient way. The integrated algorithm is formulated in a recursive Bayesian framework that sequentially segments cardiac images and estimates myocardial displacements. ldquoLeave-one-outrdquo validation on 32 sequences demonstrates that the segmentation results are improved when the SSDM is used, and the tracking results are much more accurate when the segmentation module is added.","PeriodicalId":102206,"journal":{"name":"2008 IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-06-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128319264","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2008-06-23DOI: 10.1109/CVPRW.2008.4563021
Sergio Escalera, O. Pujol, J. Mauri, P. Radeva
Intravascular ultrasound (IVUS) represents a powerful imaging technique to explore coronary vessels and to study their morphology and histologic properties. In this paper, we characterize different tissues based on Radio Frequency, texture-based, slope-based, and combined features. To deal with the classification of multiple tissues, we require the use of robust multi-class learning techniques. In this context, we propose a strategy to model multi-class classification tasks using sub-classes information in the ECOC framework. The new strategy splits the classes into different subsets according to the applied base classifier. Complex IVUS data sets containing overlapping data are learnt by splitting the original set of classes into sub-classes, and embedding the binary problems in a problem-dependent ECOC design. The method automatically characterizes different tissues, showing performance improvements over the state-of-the-art ECOC techniques for different base classifiers and feature sets.
{"title":"IVUS tissue characterization with sub-class error-correcting output codes","authors":"Sergio Escalera, O. Pujol, J. Mauri, P. Radeva","doi":"10.1109/CVPRW.2008.4563021","DOIUrl":"https://doi.org/10.1109/CVPRW.2008.4563021","url":null,"abstract":"Intravascular ultrasound (IVUS) represents a powerful imaging technique to explore coronary vessels and to study their morphology and histologic properties. In this paper, we characterize different tissues based on Radio Frequency, texture-based, slope-based, and combined features. To deal with the classification of multiple tissues, we require the use of robust multi-class learning techniques. In this context, we propose a strategy to model multi-class classification tasks using sub-classes information in the ECOC framework. The new strategy splits the classes into different subsets according to the applied base classifier. Complex IVUS data sets containing overlapping data are learnt by splitting the original set of classes into sub-classes, and embedding the binary problems in a problem-dependent ECOC design. The method automatically characterizes different tissues, showing performance improvements over the state-of-the-art ECOC techniques for different base classifiers and feature sets.","PeriodicalId":102206,"journal":{"name":"2008 IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops","volume":"66 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-06-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128613575","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2008-06-23DOI: 10.1109/CVPRW.2008.4563176
A. Yang, Sameer Iyengar, S. Sastry, R. Bajcsy, P. Kuryloski, R. Jafari
We propose a distributed recognition method to classify human actions using a low-bandwidth wearable motion sensor network. Given a set of pre-segmented motion sequences as training examples, the algorithm simultaneously segments and classifies human actions, and it also rejects outlying actions that are not in the training set. The classification is distributedly operated on individual sensor nodes and a base station computer. We show that the distribution of multiple action classes satisfies a mixture subspace model, one sub-space for each action class. Given a new test sample, we seek the sparsest linear representation of the sample w.r.t. all training examples. We show that the dominant coefficients in the representation only correspond to the action class of the test sample, and hence its membership is encoded in the representation. We further provide fast linear solvers to compute such representation via l1-minimization. Using up to eight body sensors, the algorithm achieves state-of-the-art 98.8% accuracy on a set of 12 action categories. We further demonstrate that the recognition precision only decreases gracefully using smaller subsets of sensors, which validates the robustness of the distributed framework.
{"title":"Distributed segmentation and classification of human actions using a wearable motion sensor network","authors":"A. Yang, Sameer Iyengar, S. Sastry, R. Bajcsy, P. Kuryloski, R. Jafari","doi":"10.1109/CVPRW.2008.4563176","DOIUrl":"https://doi.org/10.1109/CVPRW.2008.4563176","url":null,"abstract":"We propose a distributed recognition method to classify human actions using a low-bandwidth wearable motion sensor network. Given a set of pre-segmented motion sequences as training examples, the algorithm simultaneously segments and classifies human actions, and it also rejects outlying actions that are not in the training set. The classification is distributedly operated on individual sensor nodes and a base station computer. We show that the distribution of multiple action classes satisfies a mixture subspace model, one sub-space for each action class. Given a new test sample, we seek the sparsest linear representation of the sample w.r.t. all training examples. We show that the dominant coefficients in the representation only correspond to the action class of the test sample, and hence its membership is encoded in the representation. We further provide fast linear solvers to compute such representation via l1-minimization. Using up to eight body sensors, the algorithm achieves state-of-the-art 98.8% accuracy on a set of 12 action categories. We further demonstrate that the recognition precision only decreases gracefully using smaller subsets of sensors, which validates the robustness of the distributed framework.","PeriodicalId":102206,"journal":{"name":"2008 IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops","volume":"26 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-06-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130036868","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2008-06-23DOI: 10.1109/CVPRW.2008.4562976
T. Sawada, Z. Pizlo
We present a new computational model for verifying whether a 3D shape is mirror-symmetric based on its single 2D image. First, a psychophysical experiment which tested human performance in detection of 3D symmetry is described. These psychophysical results led to the formulation of a new algorithm for symmetry detection. The algorithm first recovers the 3D shape using a priori constraints (symmetry, planarity of contours and 3D compactness) and then evaluates the degree of symmetry of the 3D shape. Reliable discrimination by the algorithm between symmetric and asymmetric 3D shapes involves two measures: similarity of the two halves of a 3D shape and compactness of the 3D shape. Performance of this algorithm is highly correlated with that of the subjects. We conclude that this algorithm is a plausible model of the mechanisms used by the human visual system.
{"title":"Detecting mirror-symmetry of a volumetric shape from its single 2D image","authors":"T. Sawada, Z. Pizlo","doi":"10.1109/CVPRW.2008.4562976","DOIUrl":"https://doi.org/10.1109/CVPRW.2008.4562976","url":null,"abstract":"We present a new computational model for verifying whether a 3D shape is mirror-symmetric based on its single 2D image. First, a psychophysical experiment which tested human performance in detection of 3D symmetry is described. These psychophysical results led to the formulation of a new algorithm for symmetry detection. The algorithm first recovers the 3D shape using a priori constraints (symmetry, planarity of contours and 3D compactness) and then evaluates the degree of symmetry of the 3D shape. Reliable discrimination by the algorithm between symmetric and asymmetric 3D shapes involves two measures: similarity of the two halves of a 3D shape and compactness of the 3D shape. Performance of this algorithm is highly correlated with that of the subjects. We conclude that this algorithm is a plausible model of the mechanisms used by the human visual system.","PeriodicalId":102206,"journal":{"name":"2008 IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-06-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130173445","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：