Pub Date : 2014-10-01DOI: 10.1109/ICIP.2014.7025073
F. Comaschi, S. Stuijk, T. Basten, H. Corporaal
Object detection and tracking is one of the most important components in computer vision applications. To carefully evaluate the performance of detection and tracking algorithms, it is important to develop benchmark data sets. One of the most tedious and error-prone aspects when developing benchmarks, is the generation of the ground truth. This paper presents FAST-GT (FAst Semi-automatic Tool for Ground Truth generation), a new generic framework for the semiautomatic generation of ground truths. FAST-GT reduces the need for manual intervention thus speeding-up the ground-truthing process.
目标检测与跟踪是计算机视觉应用的重要组成部分之一。为了仔细评估检测和跟踪算法的性能，开发基准数据集非常重要。在开发基准时，最乏味和最容易出错的一个方面是生成基本事实。本文提出了FAst - gt (FAst Semi-automatic Tool for Ground Truth generation)，一种新的地面真理半自动生成通用框架。FAST-GT减少了人工干预的需要，从而加快了地面测实过程。
{"title":"A tool for fast ground truth generation for object detection and tracking from video","authors":"F. Comaschi, S. Stuijk, T. Basten, H. Corporaal","doi":"10.1109/ICIP.2014.7025073","DOIUrl":"https://doi.org/10.1109/ICIP.2014.7025073","url":null,"abstract":"Object detection and tracking is one of the most important components in computer vision applications. To carefully evaluate the performance of detection and tracking algorithms, it is important to develop benchmark data sets. One of the most tedious and error-prone aspects when developing benchmarks, is the generation of the ground truth. This paper presents FAST-GT (FAst Semi-automatic Tool for Ground Truth generation), a new generic framework for the semiautomatic generation of ground truths. FAST-GT reduces the need for manual intervention thus speeding-up the ground-truthing process.","PeriodicalId":6856,"journal":{"name":"2014 IEEE International Conference on Image Processing (ICIP)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2014-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"72752699","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 : 2014-10-01DOI: 10.1109/ICIP.2014.7025838
Xiaoran Cao, Yun He
Video coding standards commonly use discrete cosine transform (DCT) to transform the motion compensation (M-C) residuals. However, the MC residuals have much weaker correlation than image pixels, and DCT is not the optimized transform for them. In this paper, we propose an adaptive transform structure for MC residuals. Unlike traditional approaches which use a predefined transform core, we apply singular value decomposition (SVD) on the prediction block and use the eigenvector matrices as the transform core. Experiments show that this adaptive transform is more efficient compared with the traditional approach. An average 2.0% bit rate reduction is achieved when implemented on H.265/HEVC.
{"title":"Singular vector decomposition based adaptive transform for motion compensation residuals","authors":"Xiaoran Cao, Yun He","doi":"10.1109/ICIP.2014.7025838","DOIUrl":"https://doi.org/10.1109/ICIP.2014.7025838","url":null,"abstract":"Video coding standards commonly use discrete cosine transform (DCT) to transform the motion compensation (M-C) residuals. However, the MC residuals have much weaker correlation than image pixels, and DCT is not the optimized transform for them. In this paper, we propose an adaptive transform structure for MC residuals. Unlike traditional approaches which use a predefined transform core, we apply singular value decomposition (SVD) on the prediction block and use the eigenvector matrices as the transform core. Experiments show that this adaptive transform is more efficient compared with the traditional approach. An average 2.0% bit rate reduction is achieved when implemented on H.265/HEVC.","PeriodicalId":6856,"journal":{"name":"2014 IEEE International Conference on Image Processing (ICIP)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2014-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"73399527","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 : 2014-10-01DOI: 10.1109/ICIP.2014.7025342
L. Gharsalli, B. Duchêne, A. Mohammad-Djafari, H. Ayasso
In this paper a nonlinear inverse scattering problem is solved by means of a variational Bayesian approach. The objective is to detect breast tumor from measurements of the scattered fields at different frequencies and for several illuminations. This inverse problem is known to be non linear and ill-posed. Thus, it needs to be regularized by introducing a priori information. Herein, prior information available on the sought object is that it is composed of a finite known number of different materials distributed in compact regions. It is accounted for by tackling the problem in a Bayesian framework. Then, the true joint posterior is approximated by a separable law by mean of a gradient-like variational Bayesian technique. The latter is adapted to complex valued contrast and used to compute the posterior estimators through a joint update of the shape parameters of the approximating marginals. Both permittivity and conductivity maps are reconstructed and the results obtained on synthetic data show a good reconstruction quality and a convergence faster than that of the classical variational Bayesian approach.
{"title":"A gradient-like variational Bayesian approach: Application to microwave imaging for breast tumor detection","authors":"L. Gharsalli, B. Duchêne, A. Mohammad-Djafari, H. Ayasso","doi":"10.1109/ICIP.2014.7025342","DOIUrl":"https://doi.org/10.1109/ICIP.2014.7025342","url":null,"abstract":"In this paper a nonlinear inverse scattering problem is solved by means of a variational Bayesian approach. The objective is to detect breast tumor from measurements of the scattered fields at different frequencies and for several illuminations. This inverse problem is known to be non linear and ill-posed. Thus, it needs to be regularized by introducing a priori information. Herein, prior information available on the sought object is that it is composed of a finite known number of different materials distributed in compact regions. It is accounted for by tackling the problem in a Bayesian framework. Then, the true joint posterior is approximated by a separable law by mean of a gradient-like variational Bayesian technique. The latter is adapted to complex valued contrast and used to compute the posterior estimators through a joint update of the shape parameters of the approximating marginals. Both permittivity and conductivity maps are reconstructed and the results obtained on synthetic data show a good reconstruction quality and a convergence faster than that of the classical variational Bayesian approach.","PeriodicalId":6856,"journal":{"name":"2014 IEEE International Conference on Image Processing (ICIP)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2014-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"73412425","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 : 2014-10-01DOI: 10.1109/ICIP.2014.7026065
Xiaoxia Sun, N. Nasrabadi, T. Tran
In hyperspectral pixel classification, previous research have shown that the sparse representation classifier can achieve a better performance when exploiting the neighboring test pixels through enforcing different structured sparsity priors. In this paper, we propose a supervised sparse-representation-based dictionary learning method with joint or Laplacian s-parsity priors. The proposed method has numerous advantages over the existing dictionary learning techniques. It uses a structured sparsity and provides a more robust and stable sparse coefficients. Besides, it is capable of reducing the classification error by jointly optimizing the dictionary and the classifier's parameters during the dictionary training stage.
{"title":"Task-driven dictionary learning for hyperspectral image classification with structured sparsity priors","authors":"Xiaoxia Sun, N. Nasrabadi, T. Tran","doi":"10.1109/ICIP.2014.7026065","DOIUrl":"https://doi.org/10.1109/ICIP.2014.7026065","url":null,"abstract":"In hyperspectral pixel classification, previous research have shown that the sparse representation classifier can achieve a better performance when exploiting the neighboring test pixels through enforcing different structured sparsity priors. In this paper, we propose a supervised sparse-representation-based dictionary learning method with joint or Laplacian s-parsity priors. The proposed method has numerous advantages over the existing dictionary learning techniques. It uses a structured sparsity and provides a more robust and stable sparse coefficients. Besides, it is capable of reducing the classification error by jointly optimizing the dictionary and the classifier's parameters during the dictionary training stage.","PeriodicalId":6856,"journal":{"name":"2014 IEEE International Conference on Image Processing (ICIP)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2014-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"73420374","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 : 2014-10-01DOI: 10.1109/ICIP.2014.7025685
A. R. Khalid, Ali Hassan, M. Taj
In 2D pose estimation, each limb is parametrized by it position(2D), scale(1D) and orientation(1D). One of the key bottlenecks is the exhaustive search in this 4D limb space where only a few maxima in the space are desired. To reduce the search space, we reformulate this problem in terms of finding the modes of a likelihood distribution and solve it using the Mean-Shift algorithm. Ours is the first paper in the pose estimation community to use such an approach. In addition, we describe a complete top-down approach that estimates limbs in a sequential pair-wise manner. This allows us to use Kinematic Constraints before processing, requiring us to perform search in only a small sub-region of the image for each limb. We finally devise a PCA based pose validation criteria that enables us to prune invalid hypotheses. Combining these search-space reduction techniques allows our method to generate results at par with the state-of-the-art, while saving more than 80% computations when compared to full image search.
{"title":"Efficient 2D human pose estimation using mean-shift","authors":"A. R. Khalid, Ali Hassan, M. Taj","doi":"10.1109/ICIP.2014.7025685","DOIUrl":"https://doi.org/10.1109/ICIP.2014.7025685","url":null,"abstract":"In 2D pose estimation, each limb is parametrized by it position(2D), scale(1D) and orientation(1D). One of the key bottlenecks is the exhaustive search in this 4D limb space where only a few maxima in the space are desired. To reduce the search space, we reformulate this problem in terms of finding the modes of a likelihood distribution and solve it using the Mean-Shift algorithm. Ours is the first paper in the pose estimation community to use such an approach. In addition, we describe a complete top-down approach that estimates limbs in a sequential pair-wise manner. This allows us to use Kinematic Constraints before processing, requiring us to perform search in only a small sub-region of the image for each limb. We finally devise a PCA based pose validation criteria that enables us to prune invalid hypotheses. Combining these search-space reduction techniques allows our method to generate results at par with the state-of-the-art, while saving more than 80% computations when compared to full image search.","PeriodicalId":6856,"journal":{"name":"2014 IEEE International Conference on Image Processing (ICIP)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2014-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"80111903","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 : 2014-10-01DOI: 10.1109/ICIP.2014.7025958
Mahmoud Melkemi, M. Elbaz
The problem of reconstructing the shape of dot patterns (sampled planar connected regions) is extensively studied in the literature. Up till now, the existing works do not provide guarantee for the correctness of the obtained solution, usually the results was validated empirically according to human perception. In this article, we present a new algorithm that guarantees reconstruction of the shape for a set of points satisfying some density conditions. Many experimental results show that the algorithm usually gives an adequate reconstruction for non-uniformly and weakly-sampled patterns. An advantage of the algorithm is its simplicity. Once the Delaunay triangulation of the input data is computed, simple rules are applied to the Delaunay edges in order to select those belonging to the reconstruction graph.
{"title":"A simple and efficient algorithm for dot patterns reconstruction","authors":"Mahmoud Melkemi, M. Elbaz","doi":"10.1109/ICIP.2014.7025958","DOIUrl":"https://doi.org/10.1109/ICIP.2014.7025958","url":null,"abstract":"The problem of reconstructing the shape of dot patterns (sampled planar connected regions) is extensively studied in the literature. Up till now, the existing works do not provide guarantee for the correctness of the obtained solution, usually the results was validated empirically according to human perception. In this article, we present a new algorithm that guarantees reconstruction of the shape for a set of points satisfying some density conditions. Many experimental results show that the algorithm usually gives an adequate reconstruction for non-uniformly and weakly-sampled patterns. An advantage of the algorithm is its simplicity. Once the Delaunay triangulation of the input data is computed, simple rules are applied to the Delaunay edges in order to select those belonging to the reconstruction graph.","PeriodicalId":6856,"journal":{"name":"2014 IEEE International Conference on Image Processing (ICIP)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2014-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"81374489","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 : 2014-10-01DOI: 10.1109/ICIP.2014.7025078
Zhangjian Ji, Weiqiang Wang
Recently, sparse representation has been widely applied to some generative tracking methods, which learn the representation of each particle independently and do not consider the correlation between the representation of each particle in the time domain. In this paper, we formulate the object tracking in a particle filter framework as a multi-task dynamic sparse learning problem, which we denote as Multi-Task Dynamic Sparse Tracking(MTDST). By exploring the popular sparsity-inducing ℓ1, 2 mixed norms, we regularize the representation problem to enforce joint sparsity and learn the particle representations together. Meanwhile, we also introduce the innovation sparse term in the tracking model. As compared to previous methods, our method mines the independencies between particles and the correlation of particle representation in the time domain, which improves the tracking performance. In addition, because the loft least square is robust to the outliers, we adopt the loft least square to replace the least square to calculate the likelihood probability. In the updating scheme, we eliminate the influences of occlusion pixels when updating the templates. The comprehensive experiments on the several challenging image sequences demonstrate that the proposed method consistently outperforms the existing state-of-the-art methods.
{"title":"Robust object tracking via multi-task dynamic sparse model","authors":"Zhangjian Ji, Weiqiang Wang","doi":"10.1109/ICIP.2014.7025078","DOIUrl":"https://doi.org/10.1109/ICIP.2014.7025078","url":null,"abstract":"Recently, sparse representation has been widely applied to some generative tracking methods, which learn the representation of each particle independently and do not consider the correlation between the representation of each particle in the time domain. In this paper, we formulate the object tracking in a particle filter framework as a multi-task dynamic sparse learning problem, which we denote as Multi-Task Dynamic Sparse Tracking(MTDST). By exploring the popular sparsity-inducing ℓ1, 2 mixed norms, we regularize the representation problem to enforce joint sparsity and learn the particle representations together. Meanwhile, we also introduce the innovation sparse term in the tracking model. As compared to previous methods, our method mines the independencies between particles and the correlation of particle representation in the time domain, which improves the tracking performance. In addition, because the loft least square is robust to the outliers, we adopt the loft least square to replace the least square to calculate the likelihood probability. In the updating scheme, we eliminate the influences of occlusion pixels when updating the templates. The comprehensive experiments on the several challenging image sequences demonstrate that the proposed method consistently outperforms the existing state-of-the-art methods.","PeriodicalId":6856,"journal":{"name":"2014 IEEE International Conference on Image Processing (ICIP)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2014-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"82121908","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 : 2014-10-01DOI: 10.1109/ICIP.2014.7025686
Meltem Demirkus, Doina Precup, James J. Clark, T. Arbel
Head pose estimation has been receiving a lot of attention due to its wide range of possible applications. However, most approaches in the literature have focused on head pose estimation in controlled environments. Head pose estimation has recently begun to be applied to real-world environments. However, the focus has been on estimation from single images or video frames. Furthermore, most approaches frame the problem as classification into a set of coarse pose bins, rather than performing continuous pose estimation. The proposed multi-layer probabilistic temporal graphical model robustly estimates continuous head pose angle while leveraging the strengths of multiple features into account. Experiments performed on a large, real-world video database show that our approach not only significantly outperforms alternative head pose approaches, but also provides a pose probability assigned at each video frame, which permits robust temporal, probabilistic fusion of pose information over the entire video sequence.
{"title":"Multi-layer temporal graphical model for head pose estimation in real-world videos","authors":"Meltem Demirkus, Doina Precup, James J. Clark, T. Arbel","doi":"10.1109/ICIP.2014.7025686","DOIUrl":"https://doi.org/10.1109/ICIP.2014.7025686","url":null,"abstract":"Head pose estimation has been receiving a lot of attention due to its wide range of possible applications. However, most approaches in the literature have focused on head pose estimation in controlled environments. Head pose estimation has recently begun to be applied to real-world environments. However, the focus has been on estimation from single images or video frames. Furthermore, most approaches frame the problem as classification into a set of coarse pose bins, rather than performing continuous pose estimation. The proposed multi-layer probabilistic temporal graphical model robustly estimates continuous head pose angle while leveraging the strengths of multiple features into account. Experiments performed on a large, real-world video database show that our approach not only significantly outperforms alternative head pose approaches, but also provides a pose probability assigned at each video frame, which permits robust temporal, probabilistic fusion of pose information over the entire video sequence.","PeriodicalId":6856,"journal":{"name":"2014 IEEE International Conference on Image Processing (ICIP)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2014-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"82467481","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 : 2014-10-01DOI: 10.1109/ICIP.2014.7025700
J. Heather
A robust multi-camera calibration algorithm developed for an industrial vision system is described. An ellipsoid with a simple surface pattern and accurately known geometry is used as a calibration target. Our algorithm automatically detects the presence of the ball on the conveyor and accurately determines the position of its outline and marker lines in each image frame using efficient image processing techniques. A fast least-squares minimization is then performed to determine the optimal camera and motion parameters. The method is fully automatic and requires no human interaction or guidance, helping to minimize machine setup and maintenance times. The calibration algorithm has been demonstrated on real image captures and performance is quantified using simulated image sequences.
{"title":"Calibration of an industrial vision system using an ellipsoid","authors":"J. Heather","doi":"10.1109/ICIP.2014.7025700","DOIUrl":"https://doi.org/10.1109/ICIP.2014.7025700","url":null,"abstract":"A robust multi-camera calibration algorithm developed for an industrial vision system is described. An ellipsoid with a simple surface pattern and accurately known geometry is used as a calibration target. Our algorithm automatically detects the presence of the ball on the conveyor and accurately determines the position of its outline and marker lines in each image frame using efficient image processing techniques. A fast least-squares minimization is then performed to determine the optimal camera and motion parameters. The method is fully automatic and requires no human interaction or guidance, helping to minimize machine setup and maintenance times. The calibration algorithm has been demonstrated on real image captures and performance is quantified using simulated image sequences.","PeriodicalId":6856,"journal":{"name":"2014 IEEE International Conference on Image Processing (ICIP)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2014-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"76570603","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 : 2014-10-01DOI: 10.1109/ICIP.2014.7025296
Junhui Hou, Lap-Pui Chau, Ying He, N. Magnenat-Thalmann
In this paper, we propose a practical, elegant and effective scheme for compact mocap data representation. Guided by our analysis of the unique properties of mocap data, the input mocap sequence is optimally segmented into a set of subsequences. Then, we project the subsequences onto a pair of computational orthogonal matrices to explore strong low-rank characteristic within and among the subsequences. The experimental results show that the proposed scheme is much more effective for reducing the data size, compared with the existing techniques.
{"title":"Low-rank based compact representation of motion capture data","authors":"Junhui Hou, Lap-Pui Chau, Ying He, N. Magnenat-Thalmann","doi":"10.1109/ICIP.2014.7025296","DOIUrl":"https://doi.org/10.1109/ICIP.2014.7025296","url":null,"abstract":"In this paper, we propose a practical, elegant and effective scheme for compact mocap data representation. Guided by our analysis of the unique properties of mocap data, the input mocap sequence is optimally segmented into a set of subsequences. Then, we project the subsequences onto a pair of computational orthogonal matrices to explore strong low-rank characteristic within and among the subsequences. The experimental results show that the proposed scheme is much more effective for reducing the data size, compared with the existing techniques.","PeriodicalId":6856,"journal":{"name":"2014 IEEE International Conference on Image Processing (ICIP)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2014-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"76102861","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：