Along with the development of motion capture technique, more and more 3D motion databases become available. In this paper, a novel approach is presented for motion recognition and retrieval based on ensemble HMM (hidd...Along with the development of motion capture technique, more and more 3D motion databases become available. In this paper, a novel approach is presented for motion recognition and retrieval based on ensemble HMM (hidden Markov model) learning. Due to the high dimensionality of motion’s features, Isomap nonlinear dimension reduction is used for training data of ensemble HMM learning. For handling new motion data, Isomap is generalized based on the estimation of underlying eigen- functions. Then each action class is learned with one HMM. Since ensemble learning can effectively enhance supervised learning, ensembles of weak HMM learners are built. Experiment results showed that the approaches are effective for motion data recog- nition and retrieval.展开更多
This paper applied Maximum Entropy (ME) model to Pinyin-To-Character (PTC) conversion in-stead of Hidden Markov Model (HMM) that could not include complicated and long-distance lexical informa-tion. Two ME models were...This paper applied Maximum Entropy (ME) model to Pinyin-To-Character (PTC) conversion in-stead of Hidden Markov Model (HMM) that could not include complicated and long-distance lexical informa-tion. Two ME models were built based on simple and complex templates respectively, and the complex one gave better conversion result. Furthermore, conversion trigger pair of y A → y B cBwas proposed to extract the long-distance constrain feature from the corpus; and then Average Mutual Information (AMI) was used to se-lect conversion trigger pair features which were added to the ME model. The experiment shows that conver-sion error of the ME with conversion trigger pairs is reduced by 4% on a small training corpus, comparing with HMM smoothed by absolute smoothing.展开更多
Traditional data driven fault detection methods assume that the process operates in a single mode so that they cannot perform well in processes with multiple operating modes. To monitor multimode processes effectively...Traditional data driven fault detection methods assume that the process operates in a single mode so that they cannot perform well in processes with multiple operating modes. To monitor multimode processes effectively,this paper proposes a novel process monitoring scheme based on orthogonal nonnegative matrix factorization(ONMF) and hidden Markov model(HMM). The new clustering technique ONMF is employed to separate data from different process modes. The multiple HMMs for various operating modes lead to higher modeling accuracy.The proposed approach does not presume the distribution of data in each mode because the process uncertainty and dynamics can be well interpreted through the hidden Markov estimation. The HMM-based monitoring indication named negative log likelihood probability is utilized for fault detection. In order to assess the proposed monitoring strategy, a numerical example and the Tennessee Eastman process are used. The results demonstrate that this method provides efficient fault detection performance.展开更多
A face recognition system based on Support Vector Machine (SVM) and Hidden Markov Model (HMM) has been proposed. The powerful discriminative ability of SVM is combined with the temporal modeling ability of HMM. The ou...A face recognition system based on Support Vector Machine (SVM) and Hidden Markov Model (HMM) has been proposed. The powerful discriminative ability of SVM is combined with the temporal modeling ability of HMM. The output of SVM is moderated to be probability output, which replaces the Mixture of Gauss (MOG) in HMM. Wavelet transformation is used to extract observation vector, which reduces the data dimension and improves the robustness.The hybrid system is compared with pure HMM face recognition method based on ORL face database and Yale face database. Experiments results show that the hybrid method has better performance.展开更多
In this paper we theoretically report an unconventional quantum phase transition of a simple Lipkin- Meshkow-Glick model： an interacting collective spin system without external magnetic field. It is shown that this m...In this paper we theoretically report an unconventional quantum phase transition of a simple Lipkin- Meshkow-Glick model： an interacting collective spin system without external magnetic field. It is shown that this model with integer-spin can exhibit a flrst-order quantum phase transition between different disordered phases, and more intriguingly, possesses a hidden supersymmetry at the critical point. However, for half-integer spin we predict another flrst-order quantum phase transition between two different long-range-ordered phases with a vanishing energy gap, which is induced by the destructive topological quantum interference between the intanton and anti-instanton tunneling paths and accompanies spontaneously breaking of supersymmetry at the same critical point. We also show that, when the total spin-value varies from half-integer to integer this model can exhibit an abrupt variation of Berry phase from π to zero.展开更多
Web pre-fetching is one of the most popular strategies, which are proposed for reducing the perceived access delay and improving the service quality of web server. In this paper, we present a pre-fetching model based ...Web pre-fetching is one of the most popular strategies, which are proposed for reducing the perceived access delay and improving the service quality of web server. In this paper, we present a pre-fetching model based an the hidden Markov model, which mines the later information requirement concepts that the user's access path contains and makes semantic-based pre-fetching decisions. Experimental results show that our schcme has better predictive pre-fetching precision.展开更多
To deal with hidden nodes in ad hoc network, we take throughput as the metric to evaluate the performance of network. Firstly, we modeled the MAC layer of ad hoc network based on 802.11 DCF without the existence of hi...To deal with hidden nodes in ad hoc network, we take throughput as the metric to evaluate the performance of network. Firstly, we modeled the MAC layer of ad hoc network based on 802.11 DCF without the existence of hidden nodes. By means of the proposed model, we evaluated the throughput performance of DCF in multi-hop wireless networks. Secondly, we performed simulations to validate this model. The outcome of comparison is that there exists much difference in throughput between the model and the simulation. For reducing this difference, we modified the model by involving hidden nodes under the condition that the AODV route protocol was chosen, and analytically analyzed the influence of hidden nodes on theoretical throughput. At last, we validated the accuracy of this model by comparing the analytical results with simulation results.展开更多
The quantification of the sheltering and exposure effects of non-uniform sediments has been widely achieved through hiding function models. Big challenge exists so far in the model parameter that is highly variable an...The quantification of the sheltering and exposure effects of non-uniform sediments has been widely achieved through hiding function models. Big challenge exists so far in the model parameter that is highly variable and differs greatly between laboratory flumes and field streams. This paper presents an improved surface-based hiding fimction. The force balance for particle inception was formulated and the allocation of the overall bed shear stress into each group of sediments was mimicked. The new hiding function was examined against and agrees well with the documented field and flume data. It was shown that the hiding fimction is closely related to the relative flow depth and the reference elevation in the velocity profile in addition to the bed material gradation. The power law of velocity profile that applies to both flume flows and natural streams can link the flume and field data together. The hiding function with b = 1/6 and b = 1/2 is applicable to natural streams and laboratory flumes, respectively. The value orb = 0.263 also works well for gravel bed rivers. The range of the reference elevation, namely z0 = 0.4Dm-1.4Dm, is recommended for either the flume or field data. The new hiding function contributes to addressing clearer physical meanings and a useful perspective for further improvement.展开更多
We study the recent e±cosmic ray excess reported by DAMPE in a Hidden Valley Model with lepton-portal dark matter. We find the electron-portal can account for the excess well and satisfy the DM relic density and ...We study the recent e±cosmic ray excess reported by DAMPE in a Hidden Valley Model with lepton-portal dark matter. We find the electron-portal can account for the excess well and satisfy the DM relic density and direct detection bounds, while electron+muon/electron+muon+tau-portal suffers from strong constraints from lepton flavor violating observables, such as μ→3 e. We also discuss possible collider signatures of our model, both at the LHC and a future 100 Te V hadron collider.展开更多
Superresolution is an image processing technique that estimates an original high-resolutionimage from its low-resolution and degraded observations.In superresolution tasks,there have beenproblems regarding the computa...Superresolution is an image processing technique that estimates an original high-resolutionimage from its low-resolution and degraded observations.In superresolution tasks,there have beenproblems regarding the computational cost for the estimation of high-dimensional variables.Theseproblems are now being overcome by the recent development of fast computers and the developmentof powerful computational techniques such as variational Bayesian approximation.This paper reviewsa Bayesian treatment of the superresolution problem and presents its extensions based on hierarchicalmodeling by employing hidden variables.展开更多
基金Project supported by the National Natural Science Foundation of China (Nos. 60533090 and 60525108), the National Basic Research Program (973) of China (No. 2002CB312101), and the Science and Technology Project of Zhejiang Province (Nos. 2005C13032 and 2005C11001-05), China
文摘Along with the development of motion capture technique, more and more 3D motion databases become available. In this paper, a novel approach is presented for motion recognition and retrieval based on ensemble HMM (hidden Markov model) learning. Due to the high dimensionality of motion’s features, Isomap nonlinear dimension reduction is used for training data of ensemble HMM learning. For handling new motion data, Isomap is generalized based on the estimation of underlying eigen- functions. Then each action class is learned with one HMM. Since ensemble learning can effectively enhance supervised learning, ensembles of weak HMM learners are built. Experiment results showed that the approaches are effective for motion data recog- nition and retrieval.
基金Supported by the National Natural Science Foundation of China as key program (No.60435020) and The HighTechnology Research and Development Programme of China (2002AA117010-09).
文摘This paper applied Maximum Entropy (ME) model to Pinyin-To-Character (PTC) conversion in-stead of Hidden Markov Model (HMM) that could not include complicated and long-distance lexical informa-tion. Two ME models were built based on simple and complex templates respectively, and the complex one gave better conversion result. Furthermore, conversion trigger pair of y A → y B cBwas proposed to extract the long-distance constrain feature from the corpus; and then Average Mutual Information (AMI) was used to se-lect conversion trigger pair features which were added to the ME model. The experiment shows that conver-sion error of the ME with conversion trigger pairs is reduced by 4% on a small training corpus, comparing with HMM smoothed by absolute smoothing.
基金Supported by the National Natural Science Foundation of China(61374140,61403072)
文摘Traditional data driven fault detection methods assume that the process operates in a single mode so that they cannot perform well in processes with multiple operating modes. To monitor multimode processes effectively,this paper proposes a novel process monitoring scheme based on orthogonal nonnegative matrix factorization(ONMF) and hidden Markov model(HMM). The new clustering technique ONMF is employed to separate data from different process modes. The multiple HMMs for various operating modes lead to higher modeling accuracy.The proposed approach does not presume the distribution of data in each mode because the process uncertainty and dynamics can be well interpreted through the hidden Markov estimation. The HMM-based monitoring indication named negative log likelihood probability is utilized for fault detection. In order to assess the proposed monitoring strategy, a numerical example and the Tennessee Eastman process are used. The results demonstrate that this method provides efficient fault detection performance.
基金This project is supported by the National Natural Science Foundation of China (No. 69889050)
文摘A face recognition system based on Support Vector Machine (SVM) and Hidden Markov Model (HMM) has been proposed. The powerful discriminative ability of SVM is combined with the temporal modeling ability of HMM. The output of SVM is moderated to be probability output, which replaces the Mixture of Gauss (MOG) in HMM. Wavelet transformation is used to extract observation vector, which reduces the data dimension and improves the robustness.The hybrid system is compared with pure HMM face recognition method based on ORL face database and Yale face database. Experiments results show that the hybrid method has better performance.
基金supported by National Natural Science Foundation of China under Grant Nos.10775091 and 10704049
文摘In this paper we theoretically report an unconventional quantum phase transition of a simple Lipkin- Meshkow-Glick model： an interacting collective spin system without external magnetic field. It is shown that this model with integer-spin can exhibit a flrst-order quantum phase transition between different disordered phases, and more intriguingly, possesses a hidden supersymmetry at the critical point. However, for half-integer spin we predict another flrst-order quantum phase transition between two different long-range-ordered phases with a vanishing energy gap, which is induced by the destructive topological quantum interference between the intanton and anti-instanton tunneling paths and accompanies spontaneously breaking of supersymmetry at the same critical point. We also show that, when the total spin-value varies from half-integer to integer this model can exhibit an abrupt variation of Berry phase from π to zero.
基金The research is supported by the National Natural Science Foundation of China(No. 60082003)
文摘Web pre-fetching is one of the most popular strategies, which are proposed for reducing the perceived access delay and improving the service quality of web server. In this paper, we present a pre-fetching model based an the hidden Markov model, which mines the later information requirement concepts that the user's access path contains and makes semantic-based pre-fetching decisions. Experimental results show that our schcme has better predictive pre-fetching precision.
文摘To deal with hidden nodes in ad hoc network, we take throughput as the metric to evaluate the performance of network. Firstly, we modeled the MAC layer of ad hoc network based on 802.11 DCF without the existence of hidden nodes. By means of the proposed model, we evaluated the throughput performance of DCF in multi-hop wireless networks. Secondly, we performed simulations to validate this model. The outcome of comparison is that there exists much difference in throughput between the model and the simulation. For reducing this difference, we modified the model by involving hidden nodes under the condition that the AODV route protocol was chosen, and analytically analyzed the influence of hidden nodes on theoretical throughput. At last, we validated the accuracy of this model by comparing the analytical results with simulation results.
基金the Beijing Municipal Science&Technology Project(Grant No.Z141100003614052)the National Natural Science Foundation of China(Grants No.51525901&51379100)as well as by China Ministry of Science and Technology(Grant No.2011CB409901)
文摘The quantification of the sheltering and exposure effects of non-uniform sediments has been widely achieved through hiding function models. Big challenge exists so far in the model parameter that is highly variable and differs greatly between laboratory flumes and field streams. This paper presents an improved surface-based hiding fimction. The force balance for particle inception was formulated and the allocation of the overall bed shear stress into each group of sediments was mimicked. The new hiding function was examined against and agrees well with the documented field and flume data. It was shown that the hiding fimction is closely related to the relative flow depth and the reference elevation in the velocity profile in addition to the bed material gradation. The power law of velocity profile that applies to both flume flows and natural streams can link the flume and field data together. The hiding function with b = 1/6 and b = 1/2 is applicable to natural streams and laboratory flumes, respectively. The value orb = 0.263 also works well for gravel bed rivers. The range of the reference elevation, namely z0 = 0.4Dm-1.4Dm, is recommended for either the flume or field data. The new hiding function contributes to addressing clearer physical meanings and a useful perspective for further improvement.
基金supported by the National Natural Science Foundation of China(Grant No.11705093)the Korea Research Fellowship Program through the National Research Foundation of Korea(NRF)funded by the Ministry of Science and ICT(Grant No.2017H1D3A1A01014127)Institute for Basic Science(IBS)under the project code IBS-R018-D1(MZ)
文摘We study the recent e±cosmic ray excess reported by DAMPE in a Hidden Valley Model with lepton-portal dark matter. We find the electron-portal can account for the excess well and satisfy the DM relic density and direct detection bounds, while electron+muon/electron+muon+tau-portal suffers from strong constraints from lepton flavor violating observables, such as μ→3 e. We also discuss possible collider signatures of our model, both at the LHC and a future 100 Te V hadron collider.
文摘Superresolution is an image processing technique that estimates an original high-resolutionimage from its low-resolution and degraded observations.In superresolution tasks,there have beenproblems regarding the computational cost for the estimation of high-dimensional variables.Theseproblems are now being overcome by the recent development of fast computers and the developmentof powerful computational techniques such as variational Bayesian approximation.This paper reviewsa Bayesian treatment of the superresolution problem and presents its extensions based on hierarchicalmodeling by employing hidden variables.
