In the present paper, we propose a method for reconstructing the surfaces of objects from stereovision data. Both the fitness of stereo data to surfaces and interrelation between the surfaces are defined in the framework of a three-dimensional (3D) Markov random field (MRF) model. The surface reconstruction is accomplished by searching for the most likely state of the MRF model. An experimental result is shown for a real scene
{"title":"Surface Reconstruction from Stereovision Data Using a 3-D MRF of Discrete Object Models","authors":"H. Takizawa, Shinji Yamamoto","doi":"10.1109/ICPR.2006.1097","DOIUrl":"https://doi.org/10.1109/ICPR.2006.1097","url":null,"abstract":"In the present paper, we propose a method for reconstructing the surfaces of objects from stereovision data. Both the fitness of stereo data to surfaces and interrelation between the surfaces are defined in the framework of a three-dimensional (3D) Markov random field (MRF) model. The surface reconstruction is accomplished by searching for the most likely state of the MRF model. An experimental result is shown for a real scene","PeriodicalId":236033,"journal":{"name":"18th International Conference on Pattern Recognition (ICPR'06)","volume":"22 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2006-08-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115634494","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}
Kazi Shah Nawaz Ripon, Chi-Ho Tsang, S. Kwong, M. Ip
In this paper, we present a novel multi-objective evolutionary clustering approach using variable-length real jumping genes genetic algorithms (VRJGGA). The proposed algorithm that extends jumping genes genetic algorithm (JGGA) (Man et al., 2004) evolves near-optimal clustering solutions using multiple clustering criteria, without a-priori knowledge of the actual number of clusters. Experimental results based on several artificial and real-world data show that VRJGGA can obtain non-dominated and near-optimal clustering solutions in terms of different cluster quality measures and classification performance
提出了一种基于变长真实跳跃基因遗传算法(VRJGGA)的多目标进化聚类方法。提出的算法扩展了跳跃基因遗传算法(JGGA) (Man et al.， 2004)，使用多个聚类标准进化出接近最优的聚类解决方案，而不需要先验地了解实际聚类的数量。基于人工和真实数据的实验结果表明，VRJGGA在不同的聚类质量度量和分类性能方面都可以获得非支配和近最优聚类解
{"title":"Multi-Objective Evolutionary Clustering using Variable-Length Real Jumping Genes Genetic Algorithm","authors":"Kazi Shah Nawaz Ripon, Chi-Ho Tsang, S. Kwong, M. Ip","doi":"10.1109/ICPR.2006.827","DOIUrl":"https://doi.org/10.1109/ICPR.2006.827","url":null,"abstract":"In this paper, we present a novel multi-objective evolutionary clustering approach using variable-length real jumping genes genetic algorithms (VRJGGA). The proposed algorithm that extends jumping genes genetic algorithm (JGGA) (Man et al., 2004) evolves near-optimal clustering solutions using multiple clustering criteria, without a-priori knowledge of the actual number of clusters. Experimental results based on several artificial and real-world data show that VRJGGA can obtain non-dominated and near-optimal clustering solutions in terms of different cluster quality measures and classification performance","PeriodicalId":236033,"journal":{"name":"18th International Conference on Pattern Recognition (ICPR'06)","volume":"8 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2006-08-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115641818","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}
Face recognition from video can be improved if good face segmentation of the subject under test is achieved. Many video based face recognition rely on simple background modeling and coarse alignment strategies for segmentation. This work presents a face recognition from video framework based on using active appearance models (AAM) to achieve accurate face segmentation and consistent shape free representation across a video sequence. The segmentation provided by the AAM can be effectively normalized (morphed) to a mean shape. The resulting sub-image can then be delivered to conventional face recognition from video algorithms for robust classification. We present preliminary results on a dataset of 17 individuals and outline the problems encountered in this approach
{"title":"Face Recognition From Video using Active Appearance Model Segmentation","authors":"N. Faggian, A. Paplinski, Tat-Jun Chin","doi":"10.1109/ICPR.2006.526","DOIUrl":"https://doi.org/10.1109/ICPR.2006.526","url":null,"abstract":"Face recognition from video can be improved if good face segmentation of the subject under test is achieved. Many video based face recognition rely on simple background modeling and coarse alignment strategies for segmentation. This work presents a face recognition from video framework based on using active appearance models (AAM) to achieve accurate face segmentation and consistent shape free representation across a video sequence. The segmentation provided by the AAM can be effectively normalized (morphed) to a mean shape. The resulting sub-image can then be delivered to conventional face recognition from video algorithms for robust classification. We present preliminary results on a dataset of 17 individuals and outline the problems encountered in this approach","PeriodicalId":236033,"journal":{"name":"18th International Conference on Pattern Recognition (ICPR'06)","volume":"199 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2006-08-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115693088","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}
This paper presents a method for synchronizing multiple cameras from only the images captured by the cameras, assuming that they are not connected to an external clock signal source. It is assumed that The cameras are stationary and take the images of the same scene from various viewpoints, in which there are moving objects such as human in motion. The method uses the appearance changes in an image sequence as a temporal feature and matches two or more sequences by evaluating the correlation among their temporal features. We show through several experiments that the method shows good performance of synchronization in spite of its simplicity. We also present a method for synchronization in sub-frame accuracy and also for adaptively selecting regions suitable for deriving the temporal features. The latter resolves the difficulty with the case where some objects appear in some of the cameras and do not in the rest, which could deteriorate accuracy of the synchronization
{"title":"Video Synchronization Based on Co-occurrence of Appearance Changes in Video Sequences","authors":"Manabu Ushizaki, Takayuki Okatani, K. Deguchi","doi":"10.1109/ICPR.2006.1174","DOIUrl":"https://doi.org/10.1109/ICPR.2006.1174","url":null,"abstract":"This paper presents a method for synchronizing multiple cameras from only the images captured by the cameras, assuming that they are not connected to an external clock signal source. It is assumed that The cameras are stationary and take the images of the same scene from various viewpoints, in which there are moving objects such as human in motion. The method uses the appearance changes in an image sequence as a temporal feature and matches two or more sequences by evaluating the correlation among their temporal features. We show through several experiments that the method shows good performance of synchronization in spite of its simplicity. We also present a method for synchronization in sub-frame accuracy and also for adaptively selecting regions suitable for deriving the temporal features. The latter resolves the difficulty with the case where some objects appear in some of the cameras and do not in the rest, which could deteriorate accuracy of the synchronization","PeriodicalId":236033,"journal":{"name":"18th International Conference on Pattern Recognition (ICPR'06)","volume":"89 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2006-08-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124394262","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}
This paper describes an approach to constructing a 3D paper-made object from hand-written sketches. This approach consists of two phases. One is the phase that constitutes a crease pattern based on a sketch. Another is the phase that constructs a virtual origami model from the obtained crease pattern. The crease pattern is a set of line segments in an unfolded sheet of paper and often designed by origami design methods. When an origami model is folded from a crease pattern, faces in the crease pattern may be transformed into the same plane and some inconsistent objects may be represented. In order to construct feasible (fold-able) objects, we need to dispose the faces on the same plane consistently. Therefore, in this paper, a method for analyzing overlap order of faces based on simulated annealing is proposed. Furthermore, we show some examples of 3D objects constructed by our method. The proposed method is useful for packaging and architectural modeling
{"title":"Analysis of Overlapping Faces for Constructing Paper-made Objects from Sketches","authors":"H. Shimanuki, Jien Kato, Toyohide Watanabe","doi":"10.1109/ICPR.2006.269","DOIUrl":"https://doi.org/10.1109/ICPR.2006.269","url":null,"abstract":"This paper describes an approach to constructing a 3D paper-made object from hand-written sketches. This approach consists of two phases. One is the phase that constitutes a crease pattern based on a sketch. Another is the phase that constructs a virtual origami model from the obtained crease pattern. The crease pattern is a set of line segments in an unfolded sheet of paper and often designed by origami design methods. When an origami model is folded from a crease pattern, faces in the crease pattern may be transformed into the same plane and some inconsistent objects may be represented. In order to construct feasible (fold-able) objects, we need to dispose the faces on the same plane consistently. Therefore, in this paper, a method for analyzing overlap order of faces based on simulated annealing is proposed. Furthermore, we show some examples of 3D objects constructed by our method. The proposed method is useful for packaging and architectural modeling","PeriodicalId":236033,"journal":{"name":"18th International Conference on Pattern Recognition (ICPR'06)","volume":"22 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2006-08-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124421630","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}
Previous research indicates that automatic language identification systems based on phonotactic information produce the best results compared with other systems based on acoustic or prosodic information. This paper investigates two different approaches that use phonotactic information: parallel phoneme recognition followed by language modeling (PPRLM) and multi-lingual PRLM. In the PPRLM approach, we have modified the system by using four different language models with different discounting methods, including the linear, absolute, good-turning and Witten-Bell. Our results show that the modified PPRLM system with the Witten-Bell discounting outperforms other systems and achieves 75.5% language identification accuracy for the OGI-TS speech corpus
{"title":"Multi-lingual Phoneme Recognition and Language Identification Using Phonotactic Information","authors":"Liang Wang, E. Ambikairajah, E. Choi","doi":"10.1109/ICPR.2006.823","DOIUrl":"https://doi.org/10.1109/ICPR.2006.823","url":null,"abstract":"Previous research indicates that automatic language identification systems based on phonotactic information produce the best results compared with other systems based on acoustic or prosodic information. This paper investigates two different approaches that use phonotactic information: parallel phoneme recognition followed by language modeling (PPRLM) and multi-lingual PRLM. In the PPRLM approach, we have modified the system by using four different language models with different discounting methods, including the linear, absolute, good-turning and Witten-Bell. Our results show that the modified PPRLM system with the Witten-Bell discounting outperforms other systems and achieves 75.5% language identification accuracy for the OGI-TS speech corpus","PeriodicalId":236033,"journal":{"name":"18th International Conference on Pattern Recognition (ICPR'06)","volume":"15 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2006-08-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116714670","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}
Peng Yao, Jun Yu Li, Xueyi Ye, Zhenquan Zhuang, Bin Li
In this paper, we presented an iris recognition algorithm based on modified Log-Gabor filters. The algorithm is similar as the method proposed by Daugman in general procedure while modified Log-Gabor filters are adopted to extract the iris phase information instead of complex Gabor filters used in Daugman's method. The advantage of Log-Gabor filters over complex Gabor filters is the former are strictly bandpass filters and the latter are not. The property of strictly bandpass makes the Log-Gabor filters more suitable to extract the iris phase features regardless of the background brightness. The comparison experiments between complex Gabor filters based methods and the proposed method are also presented in this paper
{"title":"Iris Recognition Algorithm Using Modified Log-Gabor Filters","authors":"Peng Yao, Jun Yu Li, Xueyi Ye, Zhenquan Zhuang, Bin Li","doi":"10.1109/ICPR.2006.726","DOIUrl":"https://doi.org/10.1109/ICPR.2006.726","url":null,"abstract":"In this paper, we presented an iris recognition algorithm based on modified Log-Gabor filters. The algorithm is similar as the method proposed by Daugman in general procedure while modified Log-Gabor filters are adopted to extract the iris phase information instead of complex Gabor filters used in Daugman's method. The advantage of Log-Gabor filters over complex Gabor filters is the former are strictly bandpass filters and the latter are not. The property of strictly bandpass makes the Log-Gabor filters more suitable to extract the iris phase features regardless of the background brightness. The comparison experiments between complex Gabor filters based methods and the proposed method are also presented in this paper","PeriodicalId":236033,"journal":{"name":"18th International Conference on Pattern Recognition (ICPR'06)","volume":"105 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2006-08-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116755417","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}
Localizing objects in images is a difficult task and represents the first step to the solution of the object recognition problem. This paper presents a novel approach to the localization problem based on recursive neural networks (RNNs), In particular, a recursive learning paradigm is proposed to process directed acyclic graphs with labeled edges, and to realize mappings between graphs which are isomorph, i.e. that share the same topology of the links. The RNN model, that assumes a graph-based representation of images, uses a state transition function that depends on the edge labels and is independent from both the number and the order of the children of each node. Moreover, the presence of targets attached to the internal nodes guarantees a fast learning, particularly sensitive to the local features of the graph. Some preliminary experiments, carried out on artificial images created using the COIL collection, are reported, showing very promising results
{"title":"Object Localization Using Input/Output Recursive Neural Networks","authors":"M. Bianchini, Marco Maggini, L. Sarti","doi":"10.1109/ICPR.2006.880","DOIUrl":"https://doi.org/10.1109/ICPR.2006.880","url":null,"abstract":"Localizing objects in images is a difficult task and represents the first step to the solution of the object recognition problem. This paper presents a novel approach to the localization problem based on recursive neural networks (RNNs), In particular, a recursive learning paradigm is proposed to process directed acyclic graphs with labeled edges, and to realize mappings between graphs which are isomorph, i.e. that share the same topology of the links. The RNN model, that assumes a graph-based representation of images, uses a state transition function that depends on the edge labels and is independent from both the number and the order of the children of each node. Moreover, the presence of targets attached to the internal nodes guarantees a fast learning, particularly sensitive to the local features of the graph. Some preliminary experiments, carried out on artificial images created using the COIL collection, are reported, showing very promising results","PeriodicalId":236033,"journal":{"name":"18th International Conference on Pattern Recognition (ICPR'06)","volume":"191 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2006-08-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116781124","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}
This paper presents a novel approach to automatically extract a subset of shape descriptors dedicated to an application under consideration. Basic descriptors having low time processing and allowing to keep nice geometric properties were implemented. Then a model based on Choquet integral and Shapley values, are proposed to select suitable descriptors. Experimental studies using real databases attest of the robustness of our approach
{"title":"Extraction of Consistent Subsets of Descriptors using Choquet Integral","authors":"Jan Rendek, Laurent Wendling","doi":"10.1109/ICPR.2006.521","DOIUrl":"https://doi.org/10.1109/ICPR.2006.521","url":null,"abstract":"This paper presents a novel approach to automatically extract a subset of shape descriptors dedicated to an application under consideration. Basic descriptors having low time processing and allowing to keep nice geometric properties were implemented. Then a model based on Choquet integral and Shapley values, are proposed to select suitable descriptors. Experimental studies using real databases attest of the robustness of our approach","PeriodicalId":236033,"journal":{"name":"18th International Conference on Pattern Recognition (ICPR'06)","volume":"9 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2006-08-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116919400","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}
Solving non-linear distortion problems in fingerprint matching is important and still remains as a challenging topic. We have developed a new fingerprint matching method to deal with non-linear distortion problems efficiently by clustering locally matched minutiae and warping the fingerprint surface using minutiae clusters. Specifically, local invariant structures encoding the neighborhood information of each minutia are utilized in clustering the matched minutiae and then the fingerprint surface is warped to describe the deformation pattern properly. Finally, to make an additional increase in performance, the overlapped region of two fingerprints is considered in the score computation stage. Experimental results show that the proposed algorithm is performed best compared with other ones
{"title":"Fingerprint Matching Method Using Minutiae Clustering and Warping","authors":"Dongjin Kwon, I. Yun, Duck Hoon Kim, Sang Uk Lee","doi":"10.1109/ICPR.2006.570","DOIUrl":"https://doi.org/10.1109/ICPR.2006.570","url":null,"abstract":"Solving non-linear distortion problems in fingerprint matching is important and still remains as a challenging topic. We have developed a new fingerprint matching method to deal with non-linear distortion problems efficiently by clustering locally matched minutiae and warping the fingerprint surface using minutiae clusters. Specifically, local invariant structures encoding the neighborhood information of each minutia are utilized in clustering the matched minutiae and then the fingerprint surface is warped to describe the deformation pattern properly. Finally, to make an additional increase in performance, the overlapped region of two fingerprints is considered in the score computation stage. Experimental results show that the proposed algorithm is performed best compared with other ones","PeriodicalId":236033,"journal":{"name":"18th International Conference on Pattern Recognition (ICPR'06)","volume":"49 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2006-08-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116945589","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：