Due to the lack of large-scale emotion databases,it is hard to obtain comparable improvement in multimodal emotion recognition of the deep neural network by deep learning,which has made great progress in other areas.W...Due to the lack of large-scale emotion databases,it is hard to obtain comparable improvement in multimodal emotion recognition of the deep neural network by deep learning,which has made great progress in other areas.We use transfer learning to improve its performance with pretrained models on largescale data.Audio is encoded using deep speech recognition networks with 500 hours’speech and video is encoded using convolutional neural networks with over 110,000 images.The extracted audio and visual features are fed into Long Short-Term Memory to train models respectively.Logistic regression and ensemble method are performed in decision level fusion.The experiment results indicate that 1)audio features extracted from deep speech recognition networks achieve better performance than handcrafted audio features;2)the visual emotion recognition obtains better performance than audio emotion recognition;3)the ensemble method gets better performance than logistic regression and prior knowledge from micro-F1 value further improves the performance and robustness,achieving accuracy of 67.00%for“happy”,54.90%for“an?gry”,and 51.69%for“sad”.展开更多
The prediction of slope stability is considered as one of the critical concerns in geotechnical engineering.Conventional stochastic analysis with spatially variable slopes is time-consuming and highly computation-dema...The prediction of slope stability is considered as one of the critical concerns in geotechnical engineering.Conventional stochastic analysis with spatially variable slopes is time-consuming and highly computation-demanding.To assess the slope stability problems with a more desirable computational effort,many machine learning(ML)algorithms have been proposed.However,most ML-based techniques require that the training data must be in the same feature space and have the same distribution,and the model may need to be rebuilt when the spatial distribution changes.This paper presents a new ML-based algorithm,which combines the principal component analysis(PCA)-based neural network(NN)and transfer learning(TL)techniques(i.e.PCAeNNeTL)to conduct the stability analysis of slopes with different spatial distributions.The Monte Carlo coupled with finite element simulation is first conducted for data acquisition considering the spatial variability of cohesive strength or friction angle of soils from eight slopes with the same geometry.The PCA method is incorporated into the neural network algorithm(i.e.PCA-NN)to increase the computational efficiency by reducing the input variables.It is found that the PCA-NN algorithm performs well in improving the prediction of slope stability for a given slope in terms of the computational accuracy and computational effort when compared with the other two algorithms(i.e.NN and decision trees,DT).Furthermore,the PCAeNNeTL algorithm shows great potential in assessing the stability of slope even with fewer training data.展开更多
Identifying faces in non-frontal poses presents a significant challenge for face recognition(FR)systems.In this study,we delved into the impact of yaw pose variations on these systems and devised a robust method for d...Identifying faces in non-frontal poses presents a significant challenge for face recognition(FR)systems.In this study,we delved into the impact of yaw pose variations on these systems and devised a robust method for detecting faces across a wide range of angles from 0°to±90°.We initially selected the most suitable feature vector size by integrating the Dlib,FaceNet(Inception-v2),and“Support Vector Machines(SVM)”+“K-nearest neighbors(KNN)”algorithms.To train and evaluate this feature vector,we used two datasets:the“Labeled Faces in the Wild(LFW)”benchmark data and the“Robust Shape-Based FR System(RSBFRS)”real-time data,which contained face images with varying yaw poses.After selecting the best feature vector,we developed a real-time FR system to handle yaw poses.The proposed FaceNet architecture achieved recognition accuracies of 99.7%and 99.8%for the LFW and RSBFRS datasets,respectively,with 128 feature vector dimensions and minimum Euclidean distance thresholds of 0.06 and 0.12.The FaceNet+SVM and FaceNet+KNN classifiers achieved classification accuracies of 99.26%and 99.44%,respectively.The 128-dimensional embedding vector showed the highest recognition rate among all dimensions.These results demonstrate the effectiveness of our proposed approach in enhancing FR accuracy,particularly in real-world scenarios with varying yaw poses.展开更多
Artificial intelligence(AI),which has recently gained popularity,is being extensively employed in modern fault diagnostic research to preserve the reliability and productivity of machines.The effectiveness of AI is in...Artificial intelligence(AI),which has recently gained popularity,is being extensively employed in modern fault diagnostic research to preserve the reliability and productivity of machines.The effectiveness of AI is influenced by the quality of the labeled training data.However,in engineering scenarios,available data on mechanical equipment are scarce,and collecting massive amounts of well-annotated fault data to train AI models is expensive and difficult.In response to the inadequacy of training samples,a numerical simulation-based partial transfer learning method for machinery fault diagnosis is proposed.First,a suitable simulation model of critical components in a mechanical system is developed using the finite element method(FEM),and numerical simulation is performed to acquire FEM simulation samples containing different fault types.Second,several synthetic simulation samples are generated to form complete source domain training samples using a generative adversarial network.Subsequently,the partial transfer learning network is trained to extract shared fault characteristics between the simulation and measured samples in the case of class imbalance.Finally,the resulting model is used to diagnose unknown samples from real-world mechanical systems in operation.The proposed method is tested on actual fault samples of bearings and gears obtained from a public dataset and experimental test rig available in our laboratory,achieving average classification accuracy of 99.54%and 99.64%,respectively.Comparison investigations reveal that the proposed method has superior classification and generalization ability when detecting faults in real mechanical systems.展开更多
文摘Due to the lack of large-scale emotion databases,it is hard to obtain comparable improvement in multimodal emotion recognition of the deep neural network by deep learning,which has made great progress in other areas.We use transfer learning to improve its performance with pretrained models on largescale data.Audio is encoded using deep speech recognition networks with 500 hours’speech and video is encoded using convolutional neural networks with over 110,000 images.The extracted audio and visual features are fed into Long Short-Term Memory to train models respectively.Logistic regression and ensemble method are performed in decision level fusion.The experiment results indicate that 1)audio features extracted from deep speech recognition networks achieve better performance than handcrafted audio features;2)the visual emotion recognition obtains better performance than audio emotion recognition;3)the ensemble method gets better performance than logistic regression and prior knowledge from micro-F1 value further improves the performance and robustness,achieving accuracy of 67.00%for“happy”,54.90%for“an?gry”,and 51.69%for“sad”.
基金supported by the National Natural Science Foundation of China(Grant No.52008402)the Central South University autonomous exploration project(Grant No.2021zzts0790).
文摘The prediction of slope stability is considered as one of the critical concerns in geotechnical engineering.Conventional stochastic analysis with spatially variable slopes is time-consuming and highly computation-demanding.To assess the slope stability problems with a more desirable computational effort,many machine learning(ML)algorithms have been proposed.However,most ML-based techniques require that the training data must be in the same feature space and have the same distribution,and the model may need to be rebuilt when the spatial distribution changes.This paper presents a new ML-based algorithm,which combines the principal component analysis(PCA)-based neural network(NN)and transfer learning(TL)techniques(i.e.PCAeNNeTL)to conduct the stability analysis of slopes with different spatial distributions.The Monte Carlo coupled with finite element simulation is first conducted for data acquisition considering the spatial variability of cohesive strength or friction angle of soils from eight slopes with the same geometry.The PCA method is incorporated into the neural network algorithm(i.e.PCA-NN)to increase the computational efficiency by reducing the input variables.It is found that the PCA-NN algorithm performs well in improving the prediction of slope stability for a given slope in terms of the computational accuracy and computational effort when compared with the other two algorithms(i.e.NN and decision trees,DT).Furthermore,the PCAeNNeTL algorithm shows great potential in assessing the stability of slope even with fewer training data.
基金funding for the project,excluding research publication,from the Board of Research in Nuclear Sciences(BRNS)under Grant Number 59/14/05/2019/BRNS.
文摘Identifying faces in non-frontal poses presents a significant challenge for face recognition(FR)systems.In this study,we delved into the impact of yaw pose variations on these systems and devised a robust method for detecting faces across a wide range of angles from 0°to±90°.We initially selected the most suitable feature vector size by integrating the Dlib,FaceNet(Inception-v2),and“Support Vector Machines(SVM)”+“K-nearest neighbors(KNN)”algorithms.To train and evaluate this feature vector,we used two datasets:the“Labeled Faces in the Wild(LFW)”benchmark data and the“Robust Shape-Based FR System(RSBFRS)”real-time data,which contained face images with varying yaw poses.After selecting the best feature vector,we developed a real-time FR system to handle yaw poses.The proposed FaceNet architecture achieved recognition accuracies of 99.7%and 99.8%for the LFW and RSBFRS datasets,respectively,with 128 feature vector dimensions and minimum Euclidean distance thresholds of 0.06 and 0.12.The FaceNet+SVM and FaceNet+KNN classifiers achieved classification accuracies of 99.26%and 99.44%,respectively.The 128-dimensional embedding vector showed the highest recognition rate among all dimensions.These results demonstrate the effectiveness of our proposed approach in enhancing FR accuracy,particularly in real-world scenarios with varying yaw poses.
文摘针对不同型号滚动轴承监测信号之间特征分布差异大、故障数据样本少,导致轴承故障精度低的问题,提出了一种基于改进交替迁移学习的滚动轴承故障诊断算法。为了充分发挥卷积神经网络(convolutional neural network, CNN)对二维数据优秀的特征提取能力,首先将一维振动信号转化为二维图像,输入到深度卷积神经网络中学习;其次,为了减少源域与目标域数据间的特征分布差异,提出了改进的交替迁移学习(improved alternately transfer learning, IATL),通过交替计算域间的CORAL损失函数和最大均值差异(maximum mean discrepancy, MMD)损失函数,并反向传播更新各层网络权重与偏置参数,以实现变工况、跨轴承型号和小故障样本条件下轴承特征迁移适配;最后,在全连接层使用Softmax函数对目标域数据进行故障诊断。为了验证该算法的有效性,采用凯斯西储大学(Case Western Reserve University, CWRU)的滚动轴承数据集进行了迁移试验验证。结果表明,与仅计算CORAL损失函数和MMD损失函数等算法对比可知,该算法有效地减少了领域数据之间的特征分布差异,具有较高的故障分类准确率。
基金supported by the National Natural Science Foundation of China (Grant No.U1909217)the Zhejiang Natural Science Foundation of China (Grant No.LD21E050001)the Wenzhou Major Science and Technology Innovation Project of China (Grant No.ZG2020051)。
文摘Artificial intelligence(AI),which has recently gained popularity,is being extensively employed in modern fault diagnostic research to preserve the reliability and productivity of machines.The effectiveness of AI is influenced by the quality of the labeled training data.However,in engineering scenarios,available data on mechanical equipment are scarce,and collecting massive amounts of well-annotated fault data to train AI models is expensive and difficult.In response to the inadequacy of training samples,a numerical simulation-based partial transfer learning method for machinery fault diagnosis is proposed.First,a suitable simulation model of critical components in a mechanical system is developed using the finite element method(FEM),and numerical simulation is performed to acquire FEM simulation samples containing different fault types.Second,several synthetic simulation samples are generated to form complete source domain training samples using a generative adversarial network.Subsequently,the partial transfer learning network is trained to extract shared fault characteristics between the simulation and measured samples in the case of class imbalance.Finally,the resulting model is used to diagnose unknown samples from real-world mechanical systems in operation.The proposed method is tested on actual fault samples of bearings and gears obtained from a public dataset and experimental test rig available in our laboratory,achieving average classification accuracy of 99.54%and 99.64%,respectively.Comparison investigations reveal that the proposed method has superior classification and generalization ability when detecting faults in real mechanical systems.
