Pub Date : 2002-12-10DOI: 10.1109/ICPR.2002.1048485
M. Koskela, Jorma T. Laaksonen, E. Oja
The MPEG-7 standard is emerging as both a general framework for content description and a collection of specific, agreed-upon content descriptors. We have developed a neural, self-organizing technique for content-based image retrieval. In this paper we apply the visual content descriptors provided by MPEG-7 in our PicSOM system and compare our own image indexing technique with a reference method based on vector quantization. The results of our experiments show that the MPEG-7 descriptors can be used as such in the PicSOM system.
{"title":"Using MPEG-7 descriptors in image retrieval with self-organizing maps","authors":"M. Koskela, Jorma T. Laaksonen, E. Oja","doi":"10.1109/ICPR.2002.1048485","DOIUrl":"https://doi.org/10.1109/ICPR.2002.1048485","url":null,"abstract":"The MPEG-7 standard is emerging as both a general framework for content description and a collection of specific, agreed-upon content descriptors. We have developed a neural, self-organizing technique for content-based image retrieval. In this paper we apply the visual content descriptors provided by MPEG-7 in our PicSOM system and compare our own image indexing technique with a reference method based on vector quantization. The results of our experiments show that the MPEG-7 descriptors can be used as such in the PicSOM system.","PeriodicalId":159502,"journal":{"name":"Object recognition supported by user interaction for service robots","volume":"7 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2002-12-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133966430","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 : 2002-12-10DOI: 10.1109/ICPR.2002.1048231
B. Fasel
Automatic face analysis has to cope with pose and lighting variations. Especially pose variations are difficult to tackle and many face analysis methods require the use of sophisticated normalization procedures. We propose a data-driven face analysis approach that is not only capable of extracting features relevant to a given face analysis task but is also robust with regard to face location changes and scale variations. This is achieved by deploying convolutional neural networks, which are either trained for facial expression recognition or face identity recognition. Combining the outputs of these networks allows us to obtain a subject dependent or personalized recognition of facial expressions.
{"title":"Robust face analysis using convolutional neural networks","authors":"B. Fasel","doi":"10.1109/ICPR.2002.1048231","DOIUrl":"https://doi.org/10.1109/ICPR.2002.1048231","url":null,"abstract":"Automatic face analysis has to cope with pose and lighting variations. Especially pose variations are difficult to tackle and many face analysis methods require the use of sophisticated normalization procedures. We propose a data-driven face analysis approach that is not only capable of extracting features relevant to a given face analysis task but is also robust with regard to face location changes and scale variations. This is achieved by deploying convolutional neural networks, which are either trained for facial expression recognition or face identity recognition. Combining the outputs of these networks allows us to obtain a subject dependent or personalized recognition of facial expressions.","PeriodicalId":159502,"journal":{"name":"Object recognition supported by user interaction for service robots","volume":"392 3","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2002-12-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133052623","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 : 2002-12-10DOI: 10.1109/ICPR.2002.1048259
Il-Seok Oh, Jin-Seon Lee, B. Moon
This paper proposes a novel hybrid genetic algorithm for the feature selection. Local search operations used to improve chromosomes are defined and embedded in hybrid GAs. The hybridization gives two desirable effects: improving the final performance significantly and acquiring control of subset size. For the implementation reproduction by readers, we provide detailed information of GA procedure and parameter setting. Experimental results reveal that the proposed hybrid GA is superior to a classical GA and sequential search algorithms.
{"title":"Local search-embedded genetic algorithms for feature selection","authors":"Il-Seok Oh, Jin-Seon Lee, B. Moon","doi":"10.1109/ICPR.2002.1048259","DOIUrl":"https://doi.org/10.1109/ICPR.2002.1048259","url":null,"abstract":"This paper proposes a novel hybrid genetic algorithm for the feature selection. Local search operations used to improve chromosomes are defined and embedded in hybrid GAs. The hybridization gives two desirable effects: improving the final performance significantly and acquiring control of subset size. For the implementation reproduction by readers, we provide detailed information of GA procedure and parameter setting. Experimental results reveal that the proposed hybrid GA is superior to a classical GA and sequential search algorithms.","PeriodicalId":159502,"journal":{"name":"Object recognition supported by user interaction for service robots","volume":"53 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2002-12-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122487887","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 : 2002-12-10DOI: 10.1109/ICPR.2002.1048380
Thomas Zöller, L. Hermes, J. Buhmann
Unsupervised image segmentation can be formulated as a clustering problem in which pixels or small image patches are grouped together based on local feature information. In this contribution, parametric distributional clustering (PDC) is presented as a novel approach to image segmentation based on color and texture clues. The objective function of the PDC model is derived from the recently proposed Information Bottleneck framework (Tishby et al., 1999), but it can equivalently be formulated in terms of a maximum likelihood solution. Its optimization is performed by deterministic annealing. Segmentation results are shown for natural wildlife imagery.
无监督图像分割可以表述为基于局部特征信息将像素或小图像块分组在一起的聚类问题。在这篇贡献中，参数分布聚类(PDC)是一种基于颜色和纹理线索的图像分割新方法。PDC模型的目标函数来源于最近提出的信息瓶颈框架(Tishby et al.， 1999)，但它可以等效地用最大似然解来表示。采用确定性退火方法对其进行优化。自然野生动物图像的分割结果。
{"title":"Combined color and texture segmentation by parametric distributional clustering","authors":"Thomas Zöller, L. Hermes, J. Buhmann","doi":"10.1109/ICPR.2002.1048380","DOIUrl":"https://doi.org/10.1109/ICPR.2002.1048380","url":null,"abstract":"Unsupervised image segmentation can be formulated as a clustering problem in which pixels or small image patches are grouped together based on local feature information. In this contribution, parametric distributional clustering (PDC) is presented as a novel approach to image segmentation based on color and texture clues. The objective function of the PDC model is derived from the recently proposed Information Bottleneck framework (Tishby et al., 1999), but it can equivalently be formulated in terms of a maximum likelihood solution. Its optimization is performed by deterministic annealing. Segmentation results are shown for natural wildlife imagery.","PeriodicalId":159502,"journal":{"name":"Object recognition supported by user interaction for service robots","volume":"20 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2002-12-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114173675","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 : 2002-12-10DOI: 10.1109/ICPR.2002.1048230
A. Rukhin, P. Grother, P. Phillips, Stefan Leigh, A. Heckert, E. Newton
Nonparametric statistics for quantifying dependence between the output rankings of face recognition algorithms are described Analysis of the archived results of a large face recognition study shows that even the better algorithms exhibit significantly different behaviors. It is found that there is significant dependence in the rankings given by two algorithms to similar and dissimilar faces but that other samples are ranked independently. A class of functions known as copulas is used; it is shown that the correlations arise from a mixture of two copulas.
{"title":"Dependence characteristics of face recognition algorithms","authors":"A. Rukhin, P. Grother, P. Phillips, Stefan Leigh, A. Heckert, E. Newton","doi":"10.1109/ICPR.2002.1048230","DOIUrl":"https://doi.org/10.1109/ICPR.2002.1048230","url":null,"abstract":"Nonparametric statistics for quantifying dependence between the output rankings of face recognition algorithms are described Analysis of the archived results of a large face recognition study shows that even the better algorithms exhibit significantly different behaviors. It is found that there is significant dependence in the rankings given by two algorithms to similar and dissimilar faces but that other samples are ranked independently. A class of functions known as copulas is used; it is shown that the correlations arise from a mixture of two copulas.","PeriodicalId":159502,"journal":{"name":"Object recognition supported by user interaction for service robots","volume":"124 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2002-12-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116647218","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 : 2002-12-10DOI: 10.1109/ICPR.2002.1048265
D. Rivière, J. F. Mangin, Jean-Marc Martinez, F. Tupin, D. Papadopoulos-Orfanos, V. Frouin
This paper introduces an approach for handling complex labelling problems driven by local constraints. The purpose is illustrated by two applications: detection of the road network on radar satellite images, and recognition of the cortical sulci on MRI images. Features must be initially extracted from the data to build a "feature graph" with structural relations. The goal is to endow each feature with a label representing either a specific object (recognition), or a class of objects (detection). Some contextual constraints have to be respected during this labelling. They are modelled by Markovian potentials assigned to the labellings of "feature clusters". The solution of the labelling problem is the minimum of the energy defined by the sum of the local potentials. This paper develops a method for learning these local potentials using "congregation" of neural networks and supervised learning.
{"title":"Relational graph labelling using learning techniques and Markov random fields","authors":"D. Rivière, J. F. Mangin, Jean-Marc Martinez, F. Tupin, D. Papadopoulos-Orfanos, V. Frouin","doi":"10.1109/ICPR.2002.1048265","DOIUrl":"https://doi.org/10.1109/ICPR.2002.1048265","url":null,"abstract":"This paper introduces an approach for handling complex labelling problems driven by local constraints. The purpose is illustrated by two applications: detection of the road network on radar satellite images, and recognition of the cortical sulci on MRI images. Features must be initially extracted from the data to build a \"feature graph\" with structural relations. The goal is to endow each feature with a label representing either a specific object (recognition), or a class of objects (detection). Some contextual constraints have to be respected during this labelling. They are modelled by Markovian potentials assigned to the labellings of \"feature clusters\". The solution of the labelling problem is the minimum of the energy defined by the sum of the local potentials. This paper develops a method for learning these local potentials using \"congregation\" of neural networks and supervised learning.","PeriodicalId":159502,"journal":{"name":"Object recognition supported by user interaction for service robots","volume":"53 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2002-12-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129455432","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 : 2002-12-10DOI: 10.1109/ICPR.2002.1048459
C. Fookes, Bennamoun, A. Lamanna
Mutual information (MI) has shown promise as an effective stereo matching measure for images affected by radiometric distortion. This is due to the robustness of MI against changes in illumination. However MI-based approaches are particularly prone to the generation of false matches due to the small statistical power of the matching windows. The paper proposes extensions to MI-based stereo matching in order to increase the robustness of the algorithm. Firstly, prior probabilities are incorporated into the MI measure in order to considerably increase the statistical power of the matching windows. These prior probabilities, which are calculated from the global joint histogram between the stereo pair, are tuned to a two level hierarchical approach. A 2D match surface, in which the match score is computed for every possible combination of template and matching window, is also utilised. This enforces left-right consistency and uniqueness constraints. These additions to MI-based stereo matching significantly enhance the algorithm's ability to detect correct matches while decreasing computation time and improving the accuracy.
{"title":"Improved stereo image matching using mutual information and hierarchical prior probabilities","authors":"C. Fookes, Bennamoun, A. Lamanna","doi":"10.1109/ICPR.2002.1048459","DOIUrl":"https://doi.org/10.1109/ICPR.2002.1048459","url":null,"abstract":"Mutual information (MI) has shown promise as an effective stereo matching measure for images affected by radiometric distortion. This is due to the robustness of MI against changes in illumination. However MI-based approaches are particularly prone to the generation of false matches due to the small statistical power of the matching windows. The paper proposes extensions to MI-based stereo matching in order to increase the robustness of the algorithm. Firstly, prior probabilities are incorporated into the MI measure in order to considerably increase the statistical power of the matching windows. These prior probabilities, which are calculated from the global joint histogram between the stereo pair, are tuned to a two level hierarchical approach. A 2D match surface, in which the match score is computed for every possible combination of template and matching window, is also utilised. This enforces left-right consistency and uniqueness constraints. These additions to MI-based stereo matching significantly enhance the algorithm's ability to detect correct matches while decreasing computation time and improving the accuracy.","PeriodicalId":159502,"journal":{"name":"Object recognition supported by user interaction for service robots","volume":"136 1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2002-12-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131206586","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 : 2002-12-10DOI: 10.1109/ICPR.2002.1048402
E. Polat, M. Yeasin, Rajeev Sharma
We present a multiple object tracking framework that employs two common methods for tracking and image matching, namely Multiple Hypothesis Tracking (MHT) and Hausdorff image matching. We use the MHT algorithm to track image edges simultaneously. This algorithm is capable of tracking multiple edges with limited occlusions and is suitable for resolving any data association uncertainty caused by background clutter and closely-spaced edges. We use the Hausdorff matching algorithm to organize individual edges into objects given their two-dimensional models. The combined technique provides a robust probabilistic tracking framework which is capable of tracking complex objects in cluttered background in video sequences.
{"title":"Multiple complex object tracking using a combined technique","authors":"E. Polat, M. Yeasin, Rajeev Sharma","doi":"10.1109/ICPR.2002.1048402","DOIUrl":"https://doi.org/10.1109/ICPR.2002.1048402","url":null,"abstract":"We present a multiple object tracking framework that employs two common methods for tracking and image matching, namely Multiple Hypothesis Tracking (MHT) and Hausdorff image matching. We use the MHT algorithm to track image edges simultaneously. This algorithm is capable of tracking multiple edges with limited occlusions and is suitable for resolving any data association uncertainty caused by background clutter and closely-spaced edges. We use the Hausdorff matching algorithm to organize individual edges into objects given their two-dimensional models. The combined technique provides a robust probabilistic tracking framework which is capable of tracking complex objects in cluttered background in video sequences.","PeriodicalId":159502,"journal":{"name":"Object recognition supported by user interaction for service robots","volume":"103 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2002-12-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128689290","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 : 2002-12-10DOI: 10.1109/ICPR.2002.1048373
Huiguang He, Jie Tian, Jing Wang, Hong Chen, X. P. Zhang
Edge detection is the basic operation in image processing and analysis. Multiresolution sequential edge linking (MSEL) Cook and Delp (1995) has a number of advantages over other edge detection schemes, such as lower false alarm rates while maintaining full connectivity of the edge. However, it is not reasonable in the initial value selection and is time consuming. For this problem, we first use anisotropic diffusion to smooth the image while keeping the edge, and then use the feedback method to optimize the initial value. We apply our method to a medical image, and experiments show that our method is more efficient and accurate than the old MSEL.
{"title":"Improved MSEL and its medical application","authors":"Huiguang He, Jie Tian, Jing Wang, Hong Chen, X. P. Zhang","doi":"10.1109/ICPR.2002.1048373","DOIUrl":"https://doi.org/10.1109/ICPR.2002.1048373","url":null,"abstract":"Edge detection is the basic operation in image processing and analysis. Multiresolution sequential edge linking (MSEL) Cook and Delp (1995) has a number of advantages over other edge detection schemes, such as lower false alarm rates while maintaining full connectivity of the edge. However, it is not reasonable in the initial value selection and is time consuming. For this problem, we first use anisotropic diffusion to smooth the image while keeping the edge, and then use the feedback method to optimize the initial value. We apply our method to a medical image, and experiments show that our method is more efficient and accurate than the old MSEL.","PeriodicalId":159502,"journal":{"name":"Object recognition supported by user interaction for service robots","volume":"44 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2002-12-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"117125021","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 : 2002-12-10DOI: 10.1109/ICPR.2002.1048389
T. Abe, Y. Matsuzawa
To extract object regions from images, the methods using region-based active contour model (ACM) have been proposed. By controlling ACM with the statistical characteristics of the image properties, these methods effect robust region extraction. However the existing methods require redundant processing and cannot adapt to complex scene images. To overcome these problems, we propose a new method for controlling region-based ACM. In the proposed method, a definite area is set along an object boundary. This area is partitioned into several subareas, and they, are iteratively deformed to make the image properties be uniform in each subarea. As a result of this clustering on the definite area, the image properties in a necessary and sufficient area can be effectively reflected on ACM control, and efficient and accurate region extraction can be achieved.
{"title":"Clustering-based control of active contour model","authors":"T. Abe, Y. Matsuzawa","doi":"10.1109/ICPR.2002.1048389","DOIUrl":"https://doi.org/10.1109/ICPR.2002.1048389","url":null,"abstract":"To extract object regions from images, the methods using region-based active contour model (ACM) have been proposed. By controlling ACM with the statistical characteristics of the image properties, these methods effect robust region extraction. However the existing methods require redundant processing and cannot adapt to complex scene images. To overcome these problems, we propose a new method for controlling region-based ACM. In the proposed method, a definite area is set along an object boundary. This area is partitioned into several subareas, and they, are iteratively deformed to make the image properties be uniform in each subarea. As a result of this clustering on the definite area, the image properties in a necessary and sufficient area can be effectively reflected on ACM control, and efficient and accurate region extraction can be achieved.","PeriodicalId":159502,"journal":{"name":"Object recognition supported by user interaction for service robots","volume":"65 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2002-12-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125675230","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：