Acoustic features chosen from the input utterances play crucial roles in speech separation. In this paper, we propose a novel complementary feature approach that performs speech separation by combining five promising ...
详细信息
ISBN:
(纸本)9781510830028
Acoustic features chosen from the input utterances play crucial roles in speech separation. In this paper, we propose a novel complementary feature approach that performs speech separation by combining five promising features, including Gammatone filterbank power spectra(GF) and multi-resolution cochleagram(MrCG) proposedrecently especially for speech separation, as a super-vector fed into deep neural network(dNN). Additionally, based on the complementary features, we do experiments with two dNN training strategies, which are restricted Boltzmann machine(rBM) pre-training anddropout combined with rectified Linear Units(re LU), to optimize the performance of dNN. The experiment results, obtained in IEEE and TIMIT corpora using fourdifferent noises at low SNr levels of 0d B and-5d B, indicate that complementary features andrBM model improve all evaluation metrics. By contrast, dropout combined with re LU system specializes in noise suppression and objective intelligibility more.
A new propagator method for joint angle anddoppler estimation based on structured least squares(SLS)is *** defining a new propagator matrix the proposed approach reduces the loss of array *** addition,SLS which will ...
详细信息
A new propagator method for joint angle anddoppler estimation based on structured least squares(SLS)is *** defining a new propagator matrix the proposed approach reduces the loss of array *** addition,SLS which will take the specific relationship between two parts of the equation into account is utilized to enhance the estimation performance when solving the rotational invariance matrix *** results are provided verifying the effectiveness of the proposed method.
In this paper, a downlink cell-free Massive MIMO system (CF-M-MIMO-S) is considered. The CF-M-MIMO-S is a distributed M-MIMO-S, where access points (APs) with very great quantity of antennas, and a much smaller number...
详细信息
In this paper, a downlink cell-free Massive MIMO system (CF-M-MIMO-S) is considered. The CF-M-MIMO-S is a distributed M-MIMO-S, where access points (APs) with very great quantity of antennas, and a much smaller number of independent users are distributedrandomly. Firstly, an approximate expression for the capacity with perfect channel state information and conjugate beamforming scheme is derived. Secondly, an energy-efficient (EE) resource allocation strategy is advanced, which is aim to maximize system EE. Specifically, the power consumption include transmitting power, calculation power and circuit power. Simulation results indicate that the throughput of derived approximate expression is very close to theoretical value. It is also demonstrated The effectiveness of proposed algorithm and the trade-off between EE and the quantity of Aps is, meanwhile, the performance of throughput of the proposed algorithm is very well.
This paper studies the beamforming design for sum secrecy rate (SSr) maximization in the Gaussian multiple-input single-output multi-receiver wiretap channel (MISO-MrWC). The optimization problem of finding the optima...
详细信息
ISBN:
(纸本)9781479999897
This paper studies the beamforming design for sum secrecy rate (SSr) maximization in the Gaussian multiple-input single-output multi-receiver wiretap channel (MISO-MrWC). The optimization problem of finding the optimal beamforming algorithm is non-convex and intractable to solve using low-complexity methods. Motivated by the thinking of zero-forcing (ZF) and signal-to-leakage-plus-noise ratio (SLNr), we propose three low-complexity beamforming algorithms for finding a local SSr optimum. The simulation results show that the SLNr-based beamforming algorithm outperforms the other two algorithms with ZF preprocessing.
In multiuser networks, selfish users refuse to provide other users with their own energy and computing resources. In order to encourage selfish users to offer theirresources, a cooperation incentive mechanism based o...
详细信息
In multiuser networks, selfish users refuse to provide other users with their own energy and computing resources. In order to encourage selfish users to offer theirresources, a cooperation incentive mechanism based on users trust degree evaluation is proposed. The mechanism helps relays to decide whether to participate in the cooperation or not by comparing trust value. Selfish users with lower trust value will be rejected on higher probability, while cooperation can improve trust value and their own communication security. Several experiments demonstrate the effectiveness of the proposed scheme in this paper. Selfish users actively participate in the cooperation comparing situations without encouragement. At the same time, the communication quality of the users is close to the cooperation with selfless relays.
Image segmentation can be used in non-destructive testing, tracking andrecognition. Level set method for image segmentation has poor performance on efficiency. In this paper, we propose to use max-flow algorithm to o...
详细信息
Image segmentation can be used in non-destructive testing, tracking andrecognition. Level set method for image segmentation has poor performance on efficiency. In this paper, we propose to use max-flow algorithm to optimize a locally improved Chan-Vese model for image segmentation in the presence of intensity inhomogeneity. The energy function of local Chan-Vese model is introduced firstly. This model consists of global term, local term and penalty term and the local term contributes the segmentation for images with intensity inhomogeneity. Then, we convert this energy function to the frame of Graph Cut whose energy function can be efficiently minimized by max-flow algorithm. As a result, the process of optimization of local Chan-Vese model can be accelerated by using max-flow algorithm. The experiments demonstrate that the proposed method can achieve satisfactory segmentation for images with intensity inhomogeneity as well as very high efficiency.
Emotion recognition, as an important part of human-computer interaction, has been extensively researched. Various studies have already verified the relationship between emotion and the event-related potentials (ErPs)....
详细信息
Emotion recognition, as an important part of human-computer interaction, has been extensively researched. Various studies have already verified the relationship between emotion and the event-related potentials (ErPs). In this paper, a new methodology for emotion recognition is investigated by detecting single-trial ErPs related to some specific level of emotions. First, a spatial filter is constructed to estimate the ErP components. Then the most discriminative spatial and temporal features of the entire ErP waveform are extracted with lineardiscriminant analysis. The performance of this method is tested by classifying the emotional valence on three levels, the extremely negative, the moderately negative and the neutral, with the support vector machine (SVM). The result shows that the proposed method is effective.
recently, refractivity from clutter (rFC) is an important means to obtain spatial reflectivity distribution. As the input of rFC, radar sea clutter power seriously affects the inversion performance. For this problem, ...
详细信息
ISBN:
(纸本)9781538616093
recently, refractivity from clutter (rFC) is an important means to obtain spatial reflectivity distribution. As the input of rFC, radar sea clutter power seriously affects the inversion performance. For this problem, we study its distribution to get more priori information of inversion. Using this way, inversion accuracy could be improved. In this paper, seven kinds of power probability density function (PPdF) are presented. Based on these distribution models, The experimental data measured in Bohai Sea are analyzed. In addition, the goodness-of-fit tests are used to test the degree of fitness between the experimental data and the models. The results show that these radar sea clutter powerdata meet log-normal distribution or meet Weibull distribution.
A dual-banddual-polarized slot antenna for Ku-band video satellite application is proposed. It mainly consists of the ground with a wide slot and a reflector. The orthogonal modes are excited by slot-coupling feeding...
详细信息
ISBN:
(纸本)9781509047444
A dual-banddual-polarized slot antenna for Ku-band video satellite application is proposed. It mainly consists of the ground with a wide slot and a reflector. The orthogonal modes are excited by slot-coupling feeding and microstrip feeding to realize dual polarization. The dual-band characteristic is realized using the multi-resonant performance of wide slot. The impedance bandwidth for two port covers the dual wide bandwidths of 10.7-11.7GHz and 11.7-12.7GHz. The gain can be up to 8.8dBi. The proposed antenna has low profile, simple structure, good port isolation and high gain. It is a good candidate for Ku-band video satellite antenna.
Modifying flow rule attack is a type of attack against controllers in SdN, leading to severe influence on network. In this paper, we propose Mcad-SA, an aware decision-making security architecture with multi-controlle...
详细信息
ISBN:
(纸本)9781467399562
Modifying flow rule attack is a type of attack against controllers in SdN, leading to severe influence on network. In this paper, we propose Mcad-SA, an aware decision-making security architecture with multi-controller, to deal with it. It schedules the controllers in a dynamic andrandom way and adopts the vote results from the majority of controllers to determine valid flow rules. Theory analysis proves its validity.
暂无评论