The Experts below are selected from a list of 22293 Experts worldwide ranked by ideXlab platform
Sicheng Zhao - One of the best experts on this subject based on the ideXlab platform.
-
Continuous Probability Distribution Prediction of Image Emotions via Multitask Shared Sparse Regression
IEEE Transactions on Multimedia, 2017Co-Authors: Sicheng Zhao, Hongxun Yao, Yue Gao, Guiguang DingAbstract:Previous works on image emotion analysis mainly focused on predicting the dominant emotion category or the average dimension values of an image for affective image classification and regression. However, this is often insufficient in various real-world applications, as the emotions that are evoked in viewers by an image are highly subjective and different. In this paper, we propose to predict the Continuous Probability Distribution of image emotions which are represented in dimensional valence-arousal space. We carried out large-scale statistical analysis on the constructed Image-Emotion-Social-Net dataset, on which we observed that the emotion Distribution can be well-modeled by a Gaussian mixture model. This model is estimated by an expectation-maximization algorithm with specified initializations. Then, we extract commonly used emotion features at different levels for each image. Finally, we formalize the emotion Distribution prediction task as a shared sparse regression (SSR) problem and extend it to multitask settings, named multitask shared sparse regression (MTSSR), to explore the latent information between different prediction tasks. SSR and MTSSR are optimized by iteratively reweighted least squares. Experiments are conducted on the Image-Emotion-Social-Net dataset with comparisons to three alternative baselines. The quantitative results demonstrate the superiority of the proposed method.
-
predicting Continuous Probability Distribution of image emotions in valence arousal space
ACM Multimedia, 2015Co-Authors: Sicheng Zhao, Hongxun Yao, Xiaolei JiangAbstract:Previous works on image emotion analysis mainly focused on assigning a dominated emotion category or the average dimension values to an image for affective image classification and regression. However, this is often insufficient in many applications, as the emotions that are evoked in viewers by an image are highly subjective and different. In this paper, we propose to predict the Continuous Probability Distribution of dimensional image emotions represented in valence-arousal space. By the statistical analysis on the constructed Image-Emotion-Social-Net dataset, we represent the emotion Distribution as a Gaussian mixture model (GMM), which is estimated by the EM algorithm. Then we extract commonly used features of different levels for each image. Finally, we formulize the emotion Distribution prediction as a multi-task shared sparse regression (MTSSR) problem, which is optimized by iteratively reweighted least squares. Besides, we introduce three baseline algorithms. Experiments conducted on the Image-Emotion-Social-Net dataset demonstrate the superiority of the proposed method, as compared to some state-of-the-art approaches.
-
ACM Multimedia - Predicting Continuous Probability Distribution of Image Emotions in Valence-Arousal Space
Proceedings of the 23rd ACM international conference on Multimedia - MM '15, 2015Co-Authors: Sicheng Zhao, Hongxun Yao, Xiaolei JiangAbstract:Previous works on image emotion analysis mainly focused on assigning a dominated emotion category or the average dimension values to an image for affective image classification and regression. However, this is often insufficient in many applications, as the emotions that are evoked in viewers by an image are highly subjective and different. In this paper, we propose to predict the Continuous Probability Distribution of dimensional image emotions represented in valence-arousal space. By the statistical analysis on the constructed Image-Emotion-Social-Net dataset, we represent the emotion Distribution as a Gaussian mixture model (GMM), which is estimated by the EM algorithm. Then we extract commonly used features of different levels for each image. Finally, we formulize the emotion Distribution prediction as a multi-task shared sparse regression (MTSSR) problem, which is optimized by iteratively reweighted least squares. Besides, we introduce three baseline algorithms. Experiments conducted on the Image-Emotion-Social-Net dataset demonstrate the superiority of the proposed method, as compared to some state-of-the-art approaches.
John L. Markley - One of the best experts on this subject based on the ideXlab platform.
-
Continuous Probability Distribution (CUPID) analysis of potentials for internal rotations.
Journal of magnetic resonance. Series B, 1996Co-Authors: Zeljko Dzakula, William M. Westler, John L. MarkleyAbstract:Abstract The c ontin u ous p robab i lity d istribution (CUPID) approach for analyzing the rotamer populations from NMR spin–spin couplings and nuclear Overhauser enhancements [Ž. Džakula, W. M. Westler, A. S. Edison, and J. L. Markley, J. Amer. Chem. Soc. 114, 6195 (1992)] can be expanded to allow computation of the rotational potential from the Fourier coefficients of the angular Probability Distribution. This approach provides a general solution to the nonnegativity problem, which appears when lack of data causes a serious truncation in the Fourier series that defines the Probability Distribution. In favorable cases, this approach also allows thermodynamic characterization of internal rotation. Use of this extension of the CUPID method is illustrated by the analysis of internal rotations in an amino acid, two peptides, and an oligosaccharide from published experimental data. Three strategies have been devised for dealing with cases where the experimental input data do not provide enough information for complete reconstruction of the potential: (1) two-dimensional grid search for the undetermined third-order Fourier coefficients of the potential, (2) transfer of these coefficients from related model compounds, and (3) restriction of the magnitudes of the Fourier coefficients as required by the assumption of fast-exchange averaging of the input parameters. In addition, equations for translating uncertainties in experimental NMR input data into errors in calculated Continuous Probability Distributions of rotamers are presented. The dependence of errors on various features of the Distributions has been studied systematically from simulations. The results show that, typically, the confidence intervals are ±30–40° for dihedral angles and ±0.2 for rotamer populations. For χ 1 rotamers of amino acids, the analysis is most sensitive to the uncertainties in C′–H β couplings. A critical reexamination of the use of Gaussian functions to reconstruct a Probability Distribution is presented. In particular, the simplifying assumption of identical widths for all Gaussian Probability peaks has been justified by showing that it does not lead to large errors in other CUPID parameters. Finally, the angular dependences of cross-relaxation rates, their uncertainties, and the potential for their use in studying χ 1 internal rotations in amino acids are discussed.
-
Conformational Analysis of Molecules with Five-Membered Rings through NMR Determination of the Continuous Probability Distribution (CUPID) for Pseudorotation
Journal of the American Chemical Society, 1996Co-Authors: Zeljko Dzakula, And Michele L. Derider, John L. MarkleyAbstract:The Continuous Probability Distribution (CUPID) method [Džakula, Ž.; Westler, W. M.; Edison, A. S.; Markley, J. L. J. Am. Chem. Soc. 1992, 114, 6195] for conformational analysis of molecules from nuclear magnetic resonance (NMR) data has been extended to the determination of molecular conformations of five-membered rings. This approach, which should be particularly useful for studies of molecules containing pyrrolidine or furanose rings, is illustrated by the analysis of NMR data from the literature for small peptides containing l- and d-prolines, hydroxyprolines, and fluoroprolines. The CUPID approach to the analysis of five-membered rings, which takes advantage of linear regression, generally yields better fits to experimental data in a shorter time than the conventional discrete approach, which utilizes nonlinear fitting procedures. The new method proved successful in a few cases in which the conventional approach failed to produce satisfactory analysis of the data. Built-in error-propagation analysis ...
-
The "CUPID" method for calculating the Continuous Probability Distribution of rotamers from NMR data
Journal of the American Chemical Society, 1992Co-Authors: Zeljko Dzakula, William M. Westler, Arthur S. Edison, John L. MarkleyAbstract:We present a new method, Continuous Probability Distribution of rotamers (CUPID), for determining the Distribution of rotamer Probability p(X) about a dihedral angle χ. This method utilizes measured vicinal homonuclear and heteronuclear spin-spin coupling constants (J) and nuclear Overhauser enhancements (NOEs) from NMR spectra and demands no prior assumption about the conformations or the degree of flexibility across the bond
Guiguang Ding - One of the best experts on this subject based on the ideXlab platform.
-
Continuous Probability Distribution Prediction of Image Emotions via Multitask Shared Sparse Regression
IEEE Transactions on Multimedia, 2017Co-Authors: Sicheng Zhao, Hongxun Yao, Yue Gao, Guiguang DingAbstract:Previous works on image emotion analysis mainly focused on predicting the dominant emotion category or the average dimension values of an image for affective image classification and regression. However, this is often insufficient in various real-world applications, as the emotions that are evoked in viewers by an image are highly subjective and different. In this paper, we propose to predict the Continuous Probability Distribution of image emotions which are represented in dimensional valence-arousal space. We carried out large-scale statistical analysis on the constructed Image-Emotion-Social-Net dataset, on which we observed that the emotion Distribution can be well-modeled by a Gaussian mixture model. This model is estimated by an expectation-maximization algorithm with specified initializations. Then, we extract commonly used emotion features at different levels for each image. Finally, we formalize the emotion Distribution prediction task as a shared sparse regression (SSR) problem and extend it to multitask settings, named multitask shared sparse regression (MTSSR), to explore the latent information between different prediction tasks. SSR and MTSSR are optimized by iteratively reweighted least squares. Experiments are conducted on the Image-Emotion-Social-Net dataset with comparisons to three alternative baselines. The quantitative results demonstrate the superiority of the proposed method.
Hongxun Yao - One of the best experts on this subject based on the ideXlab platform.
-
Continuous Probability Distribution Prediction of Image Emotions via Multitask Shared Sparse Regression
IEEE Transactions on Multimedia, 2017Co-Authors: Sicheng Zhao, Hongxun Yao, Yue Gao, Guiguang DingAbstract:Previous works on image emotion analysis mainly focused on predicting the dominant emotion category or the average dimension values of an image for affective image classification and regression. However, this is often insufficient in various real-world applications, as the emotions that are evoked in viewers by an image are highly subjective and different. In this paper, we propose to predict the Continuous Probability Distribution of image emotions which are represented in dimensional valence-arousal space. We carried out large-scale statistical analysis on the constructed Image-Emotion-Social-Net dataset, on which we observed that the emotion Distribution can be well-modeled by a Gaussian mixture model. This model is estimated by an expectation-maximization algorithm with specified initializations. Then, we extract commonly used emotion features at different levels for each image. Finally, we formalize the emotion Distribution prediction task as a shared sparse regression (SSR) problem and extend it to multitask settings, named multitask shared sparse regression (MTSSR), to explore the latent information between different prediction tasks. SSR and MTSSR are optimized by iteratively reweighted least squares. Experiments are conducted on the Image-Emotion-Social-Net dataset with comparisons to three alternative baselines. The quantitative results demonstrate the superiority of the proposed method.
-
predicting Continuous Probability Distribution of image emotions in valence arousal space
ACM Multimedia, 2015Co-Authors: Sicheng Zhao, Hongxun Yao, Xiaolei JiangAbstract:Previous works on image emotion analysis mainly focused on assigning a dominated emotion category or the average dimension values to an image for affective image classification and regression. However, this is often insufficient in many applications, as the emotions that are evoked in viewers by an image are highly subjective and different. In this paper, we propose to predict the Continuous Probability Distribution of dimensional image emotions represented in valence-arousal space. By the statistical analysis on the constructed Image-Emotion-Social-Net dataset, we represent the emotion Distribution as a Gaussian mixture model (GMM), which is estimated by the EM algorithm. Then we extract commonly used features of different levels for each image. Finally, we formulize the emotion Distribution prediction as a multi-task shared sparse regression (MTSSR) problem, which is optimized by iteratively reweighted least squares. Besides, we introduce three baseline algorithms. Experiments conducted on the Image-Emotion-Social-Net dataset demonstrate the superiority of the proposed method, as compared to some state-of-the-art approaches.
-
ACM Multimedia - Predicting Continuous Probability Distribution of Image Emotions in Valence-Arousal Space
Proceedings of the 23rd ACM international conference on Multimedia - MM '15, 2015Co-Authors: Sicheng Zhao, Hongxun Yao, Xiaolei JiangAbstract:Previous works on image emotion analysis mainly focused on assigning a dominated emotion category or the average dimension values to an image for affective image classification and regression. However, this is often insufficient in many applications, as the emotions that are evoked in viewers by an image are highly subjective and different. In this paper, we propose to predict the Continuous Probability Distribution of dimensional image emotions represented in valence-arousal space. By the statistical analysis on the constructed Image-Emotion-Social-Net dataset, we represent the emotion Distribution as a Gaussian mixture model (GMM), which is estimated by the EM algorithm. Then we extract commonly used features of different levels for each image. Finally, we formulize the emotion Distribution prediction as a multi-task shared sparse regression (MTSSR) problem, which is optimized by iteratively reweighted least squares. Besides, we introduce three baseline algorithms. Experiments conducted on the Image-Emotion-Social-Net dataset demonstrate the superiority of the proposed method, as compared to some state-of-the-art approaches.
Xiaolei Jiang - One of the best experts on this subject based on the ideXlab platform.
-
predicting Continuous Probability Distribution of image emotions in valence arousal space
ACM Multimedia, 2015Co-Authors: Sicheng Zhao, Hongxun Yao, Xiaolei JiangAbstract:Previous works on image emotion analysis mainly focused on assigning a dominated emotion category or the average dimension values to an image for affective image classification and regression. However, this is often insufficient in many applications, as the emotions that are evoked in viewers by an image are highly subjective and different. In this paper, we propose to predict the Continuous Probability Distribution of dimensional image emotions represented in valence-arousal space. By the statistical analysis on the constructed Image-Emotion-Social-Net dataset, we represent the emotion Distribution as a Gaussian mixture model (GMM), which is estimated by the EM algorithm. Then we extract commonly used features of different levels for each image. Finally, we formulize the emotion Distribution prediction as a multi-task shared sparse regression (MTSSR) problem, which is optimized by iteratively reweighted least squares. Besides, we introduce three baseline algorithms. Experiments conducted on the Image-Emotion-Social-Net dataset demonstrate the superiority of the proposed method, as compared to some state-of-the-art approaches.
-
ACM Multimedia - Predicting Continuous Probability Distribution of Image Emotions in Valence-Arousal Space
Proceedings of the 23rd ACM international conference on Multimedia - MM '15, 2015Co-Authors: Sicheng Zhao, Hongxun Yao, Xiaolei JiangAbstract:Previous works on image emotion analysis mainly focused on assigning a dominated emotion category or the average dimension values to an image for affective image classification and regression. However, this is often insufficient in many applications, as the emotions that are evoked in viewers by an image are highly subjective and different. In this paper, we propose to predict the Continuous Probability Distribution of dimensional image emotions represented in valence-arousal space. By the statistical analysis on the constructed Image-Emotion-Social-Net dataset, we represent the emotion Distribution as a Gaussian mixture model (GMM), which is estimated by the EM algorithm. Then we extract commonly used features of different levels for each image. Finally, we formulize the emotion Distribution prediction as a multi-task shared sparse regression (MTSSR) problem, which is optimized by iteratively reweighted least squares. Besides, we introduce three baseline algorithms. Experiments conducted on the Image-Emotion-Social-Net dataset demonstrate the superiority of the proposed method, as compared to some state-of-the-art approaches.