Pub Date : 2009-06-20DOI: 10.1109/CVPRW.2009.5204334
D. Mumford
Summary form only given. Linguists have proposed dozens of formalisms for grammars and now vision is weighing in with its versions based on its needs. Ulf Grenander has proposed general pattern theory, and has used grammar-like graphical parses of "thoughts" in the style of AI. One wants a natural, simple formalism treating all these cases. I want to pose this as a central problem in modeling intelligence. Pattern theory started in the 70's with the ideas of Ulf Grenander and his school at Brown. The aim is to analyze from a statistical point of view the patterns in all "signals" generated by the world, whether they be images, sounds, written text, DNA or protein strings, spike trains in neurons, time series of prices or weather, etc. Pattern theory proposes that the types of patterns-and the hidden variables needed to describe these patterns - found in one class of signals will often be found in the others and that their characteristic variability will be similar. The underlying idea is to find classes of stochastic models which can capture all the patterns that we see in nature, so that random samples from these models have the same "look and feel" as the samples from the world itself. Then the detection of patterns in noisy and ambiguous samples can be achieved by the use of Bayes' rule, a method that can be described as "analysis by synthesis".
{"title":"Is there a general structure for grammars?","authors":"D. Mumford","doi":"10.1109/CVPRW.2009.5204334","DOIUrl":"https://doi.org/10.1109/CVPRW.2009.5204334","url":null,"abstract":"Summary form only given. Linguists have proposed dozens of formalisms for grammars and now vision is weighing in with its versions based on its needs. Ulf Grenander has proposed general pattern theory, and has used grammar-like graphical parses of \"thoughts\" in the style of AI. One wants a natural, simple formalism treating all these cases. I want to pose this as a central problem in modeling intelligence. Pattern theory started in the 70's with the ideas of Ulf Grenander and his school at Brown. The aim is to analyze from a statistical point of view the patterns in all \"signals\" generated by the world, whether they be images, sounds, written text, DNA or protein strings, spike trains in neurons, time series of prices or weather, etc. Pattern theory proposes that the types of patterns-and the hidden variables needed to describe these patterns - found in one class of signals will often be found in the others and that their characteristic variability will be similar. The underlying idea is to find classes of stochastic models which can capture all the patterns that we see in nature, so that random samples from these models have the same \"look and feel\" as the samples from the world itself. Then the detection of patterns in noisy and ambiguous samples can be achieved by the use of Bayes' rule, a method that can be described as \"analysis by synthesis\".","PeriodicalId":431981,"journal":{"name":"2009 IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops","volume":"119 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-06-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123248761","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 : 2009-06-20DOI: 10.1109/CVPRW.2009.5204044
P. MomayyezSiahkal, Kaleem Siddiqi
We approach the problem of fiber tractography from the viewpoint that a computational theory should relate to the underlying quantity that is being measured - the diffusion of water molecules. We characterize the Brownian motion of water by a 3D random walk described by a stochastic non-linear differential equation. We show that the maximum-likelihood trajectories are 3D elastica, or curves of least energy. We illustrate the model with Monte-Carlo (sequential) simulations and then develop a more efficient (local, parallelizable) implementation, based on the Fokker-Planck equation. The final algorithm allows us to efficiently compute stochastic completion fields to connect a source region to a sink region, while taking into account the underlying diffusion MRI data. We demonstrate promising tractography results using high angular resolution diffusion data as input.
{"title":"3D stochastic completion fields for fiber tractography","authors":"P. MomayyezSiahkal, Kaleem Siddiqi","doi":"10.1109/CVPRW.2009.5204044","DOIUrl":"https://doi.org/10.1109/CVPRW.2009.5204044","url":null,"abstract":"We approach the problem of fiber tractography from the viewpoint that a computational theory should relate to the underlying quantity that is being measured - the diffusion of water molecules. We characterize the Brownian motion of water by a 3D random walk described by a stochastic non-linear differential equation. We show that the maximum-likelihood trajectories are 3D elastica, or curves of least energy. We illustrate the model with Monte-Carlo (sequential) simulations and then develop a more efficient (local, parallelizable) implementation, based on the Fokker-Planck equation. The final algorithm allows us to efficiently compute stochastic completion fields to connect a source region to a sink region, while taking into account the underlying diffusion MRI data. We demonstrate promising tractography results using high angular resolution diffusion data as input.","PeriodicalId":431981,"journal":{"name":"2009 IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops","volume":"21 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-06-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123565647","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 : 2009-06-20DOI: 10.1109/CVPRW.2009.5204207
H. Seo, P. Milanfar
We present a novel bottom-up saliency detection algorithm. Our method computes so-called local regression kernels (i.e., local features) from the given image, which measure the likeness of a pixel to its surroundings. Visual saliency is then computed using the said “self-resemblance” measure. The framework results in a saliency map where each pixel indicates the statistical likelihood of saliency of a feature matrix given its surrounding feature matrices. As a similarity measure, matrix cosine similarity (a generalization of cosine similarity) is employed. State of the art performance is demonstrated on commonly used human eye fixation data [3] and some psychological patterns.
{"title":"Nonparametric bottom-up saliency detection by self-resemblance","authors":"H. Seo, P. Milanfar","doi":"10.1109/CVPRW.2009.5204207","DOIUrl":"https://doi.org/10.1109/CVPRW.2009.5204207","url":null,"abstract":"We present a novel bottom-up saliency detection algorithm. Our method computes so-called local regression kernels (i.e., local features) from the given image, which measure the likeness of a pixel to its surroundings. Visual saliency is then computed using the said “self-resemblance” measure. The framework results in a saliency map where each pixel indicates the statistical likelihood of saliency of a feature matrix given its surrounding feature matrices. As a similarity measure, matrix cosine similarity (a generalization of cosine similarity) is employed. State of the art performance is demonstrated on commonly used human eye fixation data [3] and some psychological patterns.","PeriodicalId":431981,"journal":{"name":"2009 IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops","volume":"30 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-06-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121938672","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 : 2009-06-20DOI: 10.1109/CVPRW.2009.5204274
Oksana Yakhnenko, Vasant G Honavar
Image annotation is a challenging task that allows to correlate text keywords with an image. In this paper we address the problem of image annotation using Kernel Multiple Linear Regression model. Multiple Linear Regression (MLR) model reconstructs image caption from an image by performing a linear transformation of an image into some semantic space, and then recovers the caption by performing another linear transformation from the semantic space into the label space. The model is trained so that model parameters minimize the error of reconstruction directly. This model is related to Canonical Correlation Analysis (CCA) which maps both images and caption into the semantic space to minimize the distance of mapping in the semantic space. Kernel trick is then used for the MLR resulting in Kernel Multiple Linear Regression model. The solution to KMLR is a solution to the generalized eigen-value problem, related to KCCA (Kernel Canonical Correlation Analysis). We then extend Kernel Multiple Linear Regression and Kernel Canonical Correlation analysis models to multiple kernel setting, to allow various representations of images and captions. We present results for image annotation using Multiple Kernel Learning CCA and MLR on Oliva and Torralba (2001) scene recognition that show kernel selection behaviour.
图像注释是一项具有挑战性的任务，它允许将文本关键字与图像关联起来。本文利用核多元线性回归模型解决了图像标注问题。多元线性回归(Multiple Linear Regression, MLR)模型通过对图像进行某种语义空间的线性变换来重建图像标题，然后对图像进行另一次从语义空间到标签空间的线性变换来恢复图像标题。对模型进行训练，使模型参数直接减小重构误差。该模型与典型相关分析(CCA)有关，典型相关分析将图像和标题都映射到语义空间中，以最小化语义空间中的映射距离。然后将核技巧用于MLR，从而得到核多元线性回归模型。KMLR的解是广义特征值问题的解，与核典型相关分析(KCCA)有关。然后，我们将核多元线性回归和核典型相关分析模型扩展到多个核设置，以允许图像和标题的各种表示。我们展示了在Oliva和Torralba(2001)场景识别上使用多核学习CCA和MLR进行图像注释的结果，这些结果显示了核选择行为。
{"title":"Multiple label prediction for image annotation with multiple Kernel correlation models","authors":"Oksana Yakhnenko, Vasant G Honavar","doi":"10.1109/CVPRW.2009.5204274","DOIUrl":"https://doi.org/10.1109/CVPRW.2009.5204274","url":null,"abstract":"Image annotation is a challenging task that allows to correlate text keywords with an image. In this paper we address the problem of image annotation using Kernel Multiple Linear Regression model. Multiple Linear Regression (MLR) model reconstructs image caption from an image by performing a linear transformation of an image into some semantic space, and then recovers the caption by performing another linear transformation from the semantic space into the label space. The model is trained so that model parameters minimize the error of reconstruction directly. This model is related to Canonical Correlation Analysis (CCA) which maps both images and caption into the semantic space to minimize the distance of mapping in the semantic space. Kernel trick is then used for the MLR resulting in Kernel Multiple Linear Regression model. The solution to KMLR is a solution to the generalized eigen-value problem, related to KCCA (Kernel Canonical Correlation Analysis). We then extend Kernel Multiple Linear Regression and Kernel Canonical Correlation analysis models to multiple kernel setting, to allow various representations of images and captions. We present results for image annotation using Multiple Kernel Learning CCA and MLR on Oliva and Torralba (2001) scene recognition that show kernel selection behaviour.","PeriodicalId":431981,"journal":{"name":"2009 IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops","volume":"55 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-06-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130887149","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 : 2009-06-20DOI: 10.1109/CVPRW.2009.5204309
N. Schmid, Francesco Nicolo
In the field of biometrics evaluation of quality of biometric samples has a number of important applications. The main applications include (1) to reject poor quality images during acquisition, (2) to use as enhancement metric, and (3) to apply as a weighting factor in fusion schemes. Since a biometric-based recognition system relies on measures of performance such as matching scores and recognition probability of error, it becomes intuitive that the metrics evaluating biometric sample quality have to be linked to the recognition performance of the system. The goal of this work is to design a method for evaluating and ranking various quality metrics applied to biometric images or signals based on their ability to predict recognition performance of a biometric recognition system. The proposed method involves: (1) Preprocessing algorithm operating on pairs of quality scores and generating relative scores, (2) Adaptive multivariate mapping relating quality scores and measures of recognition performance and (3) Ranking algorithm that selects the best combinations of quality measures. The performance of the method is demonstrated on face and iris biometric data.
{"title":"A method for selecting and ranking quality metrics for optimization of biometric recognition systems","authors":"N. Schmid, Francesco Nicolo","doi":"10.1109/CVPRW.2009.5204309","DOIUrl":"https://doi.org/10.1109/CVPRW.2009.5204309","url":null,"abstract":"In the field of biometrics evaluation of quality of biometric samples has a number of important applications. The main applications include (1) to reject poor quality images during acquisition, (2) to use as enhancement metric, and (3) to apply as a weighting factor in fusion schemes. Since a biometric-based recognition system relies on measures of performance such as matching scores and recognition probability of error, it becomes intuitive that the metrics evaluating biometric sample quality have to be linked to the recognition performance of the system. The goal of this work is to design a method for evaluating and ranking various quality metrics applied to biometric images or signals based on their ability to predict recognition performance of a biometric recognition system. The proposed method involves: (1) Preprocessing algorithm operating on pairs of quality scores and generating relative scores, (2) Adaptive multivariate mapping relating quality scores and measures of recognition performance and (3) Ranking algorithm that selects the best combinations of quality measures. The performance of the method is demonstrated on face and iris biometric data.","PeriodicalId":431981,"journal":{"name":"2009 IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-06-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128244961","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 : 2009-06-20DOI: 10.1109/CVPRW.2009.5204043
Xiao Han, L. Hibbard, V. Willcut
Brain structure segmentation is an important task in many neuroscience and clinical applications. In this paper, we introduce a novel MI-based dense deformable registration method and apply it to the automatic segmentation of detailed brain structures. Together with a multiple atlas fusion strategy, very accurate segmentation results were obtained, as compared with other reported methods in the literature. To make multi-atlas segmentation computationally feasible, we also propose to take advantage of the recent advancements in GPU technology and introduce a GPU-based implementation of the proposed registration method. With GPU acceleration it takes less than 8 minutes to compile a multi-atlas segmentation for each subject even with as many as 17 atlases, which demonstrates that the use of GPUs can greatly facilitate the application of such atlas-based segmentation methods in practice.
{"title":"GPU-accelerated, gradient-free MI deformable registration for atlas-based MR brain image segmentation","authors":"Xiao Han, L. Hibbard, V. Willcut","doi":"10.1109/CVPRW.2009.5204043","DOIUrl":"https://doi.org/10.1109/CVPRW.2009.5204043","url":null,"abstract":"Brain structure segmentation is an important task in many neuroscience and clinical applications. In this paper, we introduce a novel MI-based dense deformable registration method and apply it to the automatic segmentation of detailed brain structures. Together with a multiple atlas fusion strategy, very accurate segmentation results were obtained, as compared with other reported methods in the literature. To make multi-atlas segmentation computationally feasible, we also propose to take advantage of the recent advancements in GPU technology and introduce a GPU-based implementation of the proposed registration method. With GPU acceleration it takes less than 8 minutes to compile a multi-atlas segmentation for each subject even with as many as 17 atlases, which demonstrates that the use of GPUs can greatly facilitate the application of such atlas-based segmentation methods in practice.","PeriodicalId":431981,"journal":{"name":"2009 IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops","volume":"68 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-06-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114562914","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 : 2009-06-20DOI: 10.1109/CVPRW.2009.5204109
Fida El Baf, T. Bouwmans, B. Vachon
Mixture of Gaussians (MOG) is the most popular technique for background modeling and presents some limitations when dynamic changes occur in the scene like camera jitter and movement in the background. Furthermore, the MOG is initialized using a training sequence which may be noisy and/or insufficient to model correctly the background. All these critical situations generate false classification in the foreground detection mask due to the related uncertainty. In this context, we present a background modeling algorithm based on Type-2 Fuzzy Mixture of Gaussians which is particularly suitable for infrared videos. The use of the Type-2 Fuzzy Set Theory allows to take into account the uncertainty. The results using the OTCBVS benchmark/test dataset videos show the robustness of the proposed method in presence of dynamic backgrounds.
{"title":"Fuzzy statistical modeling of dynamic backgrounds for moving object detection in infrared videos","authors":"Fida El Baf, T. Bouwmans, B. Vachon","doi":"10.1109/CVPRW.2009.5204109","DOIUrl":"https://doi.org/10.1109/CVPRW.2009.5204109","url":null,"abstract":"Mixture of Gaussians (MOG) is the most popular technique for background modeling and presents some limitations when dynamic changes occur in the scene like camera jitter and movement in the background. Furthermore, the MOG is initialized using a training sequence which may be noisy and/or insufficient to model correctly the background. All these critical situations generate false classification in the foreground detection mask due to the related uncertainty. In this context, we present a background modeling algorithm based on Type-2 Fuzzy Mixture of Gaussians which is particularly suitable for infrared videos. The use of the Type-2 Fuzzy Set Theory allows to take into account the uncertainty. The results using the OTCBVS benchmark/test dataset videos show the robustness of the proposed method in presence of dynamic backgrounds.","PeriodicalId":431981,"journal":{"name":"2009 IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops","volume":"21 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-06-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114512789","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 : 2009-06-20DOI: 10.1109/CVPRW.2009.5204259
M. Mahoor, S. Cadavid, D. Messinger, J. Cohn
This paper presents a framework to automatically measure the intensity of naturally occurring facial actions. Naturalistic expressions are non-posed spontaneous actions. The facial action coding system (FACS) is the gold standard technique for describing facial expressions, which are parsed as comprehensive, nonoverlapping action units (Aus). AUs have intensities ranging from absent to maximal on a six-point metric (i.e., 0 to 5). Despite the efforts in recognizing the presence of non-posed action units, measuring their intensity has not been studied comprehensively. In this paper, we develop a framework to measure the intensity of AU12 (lip corner puller) and AU6 (cheek raising) in videos captured from infant-mother live face-to-face communications. The AU12 and AU6 are the most challenging case of infant's expressions (e.g., low facial texture in infant's face). One of the problems in facial image analysis is the large dimensionality of the visual data. Our approach for solving this problem is to utilize the spectral regression technique to project high dimensionality facial images into a low dimensionality space. Represented facial images in the low dimensional space are utilized to train support vector machine classifiers to predict the intensity of action units. Analysis of 18 minutes of captured video of non-posed facial expressions of several infants and mothers shows significant agreement between a human FACS coder and our approach, which makes it an efficient approach for automated measurement of the intensity of non-posed facial action units.
{"title":"A framework for automated measurement of the intensity of non-posed Facial Action Units","authors":"M. Mahoor, S. Cadavid, D. Messinger, J. Cohn","doi":"10.1109/CVPRW.2009.5204259","DOIUrl":"https://doi.org/10.1109/CVPRW.2009.5204259","url":null,"abstract":"This paper presents a framework to automatically measure the intensity of naturally occurring facial actions. Naturalistic expressions are non-posed spontaneous actions. The facial action coding system (FACS) is the gold standard technique for describing facial expressions, which are parsed as comprehensive, nonoverlapping action units (Aus). AUs have intensities ranging from absent to maximal on a six-point metric (i.e., 0 to 5). Despite the efforts in recognizing the presence of non-posed action units, measuring their intensity has not been studied comprehensively. In this paper, we develop a framework to measure the intensity of AU12 (lip corner puller) and AU6 (cheek raising) in videos captured from infant-mother live face-to-face communications. The AU12 and AU6 are the most challenging case of infant's expressions (e.g., low facial texture in infant's face). One of the problems in facial image analysis is the large dimensionality of the visual data. Our approach for solving this problem is to utilize the spectral regression technique to project high dimensionality facial images into a low dimensionality space. Represented facial images in the low dimensional space are utilized to train support vector machine classifiers to predict the intensity of action units. Analysis of 18 minutes of captured video of non-posed facial expressions of several infants and mothers shows significant agreement between a human FACS coder and our approach, which makes it an efficient approach for automated measurement of the intensity of non-posed facial action units.","PeriodicalId":431981,"journal":{"name":"2009 IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops","volume":"55 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-06-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126236149","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 : 2009-06-20DOI: 10.1109/CVPRW.2009.5204336
Iasonas Kokkinos, A. Yuille
Summary form only given: In this work we consider the problem of object parsing, namely detecting an object and its components by composing them from image observations. We build to address the computational complexity of the inference problem. For this we exploit our hierarchical object representation to efficiently compute a coarse solution to the problem, which we then use to guide search at a finer level. Starting from our adaptation of the A* parsing algorithm to the problem of object parsing, we then propose a coarse-to-fine approach that is capable of detecting multiple objects simultaneously. We extend this work to automatically learn a hierarchical model for a category from a set of training images for which only the bounding box is available. Our approach consists in (a) automatically registering a set of training images and constructing an object template (b) recovering object contours (c) finding object parts based on contour affinities and (d) discriminatively learning a parsing cost function.
{"title":"Inference and learning with hierarchical compositional models","authors":"Iasonas Kokkinos, A. Yuille","doi":"10.1109/CVPRW.2009.5204336","DOIUrl":"https://doi.org/10.1109/CVPRW.2009.5204336","url":null,"abstract":"Summary form only given: In this work we consider the problem of object parsing, namely detecting an object and its components by composing them from image observations. We build to address the computational complexity of the inference problem. For this we exploit our hierarchical object representation to efficiently compute a coarse solution to the problem, which we then use to guide search at a finer level. Starting from our adaptation of the A* parsing algorithm to the problem of object parsing, we then propose a coarse-to-fine approach that is capable of detecting multiple objects simultaneously. We extend this work to automatically learn a hierarchical model for a category from a set of training images for which only the bounding box is available. Our approach consists in (a) automatically registering a set of training images and constructing an object template (b) recovering object contours (c) finding object parts based on contour affinities and (d) discriminatively learning a parsing cost function.","PeriodicalId":431981,"journal":{"name":"2009 IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops","volume":"14 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-06-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128058850","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 : 2009-06-20DOI: 10.1109/CVPRW.2009.5204126
P. Khuwuthyakorn, A. Robles-Kelly, J. Zhou
In this paper, we address the problem of recovering a hyperspectral texture descriptor. We do this by viewing the wavelength-indexed bands corresponding to the texture in the image as those arising from a stochastic process whose statistics can be captured making use of the relationships between moment generating functions and Fourier kernels. In this manner, we can interpret the probability distribution of the hyper-spectral texture as a heavy-tailed one which can be rendered invariant to affine geometric transformations on the texture plane making use of the spectral power of its Fourier cosine transform. We do this by recovering the affine geometric distortion matrices corresponding to the probability density function for the texture under study. This treatment permits the development of a robust descriptor which has a high information compaction property and can capture the space and wavelength correlation for the spectra in the hyperspectral images. We illustrate the utility of our descriptor for purposes of recognition and provide results on real-world datasets. We also compare our results to those yielded by a number of alternatives.
{"title":"An affine Invariant hyperspectral texture descriptor based upon heavy-tailed distributions and fourier analysis","authors":"P. Khuwuthyakorn, A. Robles-Kelly, J. Zhou","doi":"10.1109/CVPRW.2009.5204126","DOIUrl":"https://doi.org/10.1109/CVPRW.2009.5204126","url":null,"abstract":"In this paper, we address the problem of recovering a hyperspectral texture descriptor. We do this by viewing the wavelength-indexed bands corresponding to the texture in the image as those arising from a stochastic process whose statistics can be captured making use of the relationships between moment generating functions and Fourier kernels. In this manner, we can interpret the probability distribution of the hyper-spectral texture as a heavy-tailed one which can be rendered invariant to affine geometric transformations on the texture plane making use of the spectral power of its Fourier cosine transform. We do this by recovering the affine geometric distortion matrices corresponding to the probability density function for the texture under study. This treatment permits the development of a robust descriptor which has a high information compaction property and can capture the space and wavelength correlation for the spectra in the hyperspectral images. We illustrate the utility of our descriptor for purposes of recognition and provide results on real-world datasets. We also compare our results to those yielded by a number of alternatives.","PeriodicalId":431981,"journal":{"name":"2009 IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops","volume":"51 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-06-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126655433","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：