This paper presents a local descriptor coding scheme for multicamera surveillance and 3D reconstruction embedding an autoencoder into a traditional distributed source coding strategy. The proposed solution permits shi...
详细信息
This paper presents a local descriptor coding scheme for multicamera surveillance and 3D reconstruction embedding an autoencoder into a traditional distributed source coding strategy. The proposed solution permits shifting most of the computational complexity at the decoder/receiver and exploiting the correlation among descriptors of different cameras (thus reducing the coded bit rate) without increasing the inter-device communication load. Experimental results show that the proposed scheme permits obtaining a satisfying accuracy with respect to the most recent solutions while generating a limited bit rate. (c) 2021 Elsevier B.V. All rights reserved.
Vibrotactile signals contain rich haptic information about textured surfaces but their large data volume makes it a challenging task to transmit such signals to remote locations to create immersive and realistic user ...
详细信息
ISBN:
(纸本)9781728176055
Vibrotactile signals contain rich haptic information about textured surfaces but their large data volume makes it a challenging task to transmit such signals to remote locations to create immersive and realistic user experiences. Inspired by the recent success of deep neural network (DNN) based autoencoder, we make the first attempt to apply autoencoder for lossy compression of haptic vibrotactile signals, where a convolutional neural network (CNN) and a rate-distortion (RD) function are used as the transform and cost functions, respectively. Performance comparisons with state-of-the-art methods using both peak signal-to-noise ratio (PSNR) and perceptually motivated spectral temporal similarity (ST-SIM) measures show that the proposed end-to-end vibrotactile autoencoder (EVA) is highly competitive at preserving signal quality while keeping the data rate low.
Despite advancements in brain-computer interface (BCI) research, effectively integrating electroencephalogram (EEG) and functional near-infrared spectroscopy (fNIRS) data remains a challenge due to the reliance on han...
详细信息
Despite advancements in brain-computer interface (BCI) research, effectively integrating electroencephalogram (EEG) and functional near-infrared spectroscopy (fNIRS) data remains a challenge due to the reliance on handcrafted features and limited channel selection. To address these issues, this study introduces the autoencoder-led multimodal fusion network (AMFN) for EEG–fNIRS classification. AMFN utilizes an autoencoder for automated feature extraction from EEG data and integrates these features with fNIRS data using advanced fusion techniques. This approach significantly enhances classification accuracy, surpassing traditional methods. Experimental results on motor imagery (MI) tasks demonstrate AMFN's superior performance, achieving an average intra-subject accuracy of 95.69% in left and right MI classification. This research paves the way for more intuitive and reliable BCI systems, with potential applications in robotic control, neurorehabilitation, and human-machine interaction.
This paper provides a new avenue for exploiting deep neural networks to improve physics-based simulation. Specifically, we integrate the classic Lagrangian mechanics with a deep autoencoder to accelerate elastic simul...
详细信息
This paper provides a new avenue for exploiting deep neural networks to improve physics-based simulation. Specifically, we integrate the classic Lagrangian mechanics with a deep autoencoder to accelerate elastic simulation of deformable solids. Due to the inertia effect, the dynamic equilibrium cannot be established without evaluating the second-order derivatives of the deep autoencoder network. This is beyond the capability of off-the-shelf automatic differentiation packages and algorithms, which mainly focus on the gradient evaluation. Solving the nonlinear force equilibrium is even more challenging if the standard Newton's method is to be used. This is because we need to compute a third-order derivative of the network to obtain the variational Hessian. We attack those difficulties by exploiting complex-step finite difference, coupled with reverse automatic differentiation. This strategy allows us to enjoy the convenience and accuracy of complex-step finite difference and in the meantime, to deploy complex-value perturbations as collectively as possible to save excessive network passes. With a GPU-based implementation, we are able to wield deep autoencoders (e.g., 10+ layers) with a relatively high-dimension latent space in real-time. Along this pipeline, we also design a sampling network and a weighting network to enable weight-varying Cubature integration in order to incorporate nonlinearity in the model reduction. We believe this work will inspire and benefit future research efforts in nonlinearly reduced physical simulation problems.
Seismic inversion problems often involve strong nonlinear relationships between model and data so that their misfit functions usually have many local minima. Global optimization methods are well known to be able to fi...
详细信息
Seismic inversion problems often involve strong nonlinear relationships between model and data so that their misfit functions usually have many local minima. Global optimization methods are well known to be able to find the global minimum without requiring an accurate initial model. However, when the dimensionality of model space becomes large, global optimization methods will converge slow, which seriously hinders their applications in large-dimensional seismic inversion problems. In this article, we propose a new method for large-dimensional seismic inversion based on global optimization and a machine learning technique called autoencoder. Benefiting from the dimensionality reduction characteristics of autoencoder, the proposed method converts the original large-dimensional seismic inversion problem into a low-dimensional one that can be effectively and efficiently solved by global optimization. We apply the proposed method to seismic impedance inversion problems to test its performance. We use a trace-by-trace inversion strategy, and regularization is used to guarantee the lateral continuity of the inverted model. Well-log data with accurate velocity and density are the prerequisite of the inversion strategy to work effectively. Numerical results of both synthetic and field data examples clearly demonstrate that the proposed method can converge faster and yield better inversion results compared with common methods.
Early detection, early diagnosis and classification of the cancer type facilitates faster disease management of patients. Cervical cancer is fourth most pervasive cancer type which affects life of many people worldwid...
详细信息
Early detection, early diagnosis and classification of the cancer type facilitates faster disease management of patients. Cervical cancer is fourth most pervasive cancer type which affects life of many people worldwide. The intent of this study is to automate cancer diagnosis and classification through deep learning techniques to ensure patients health condition progress timely. For this research, Herlev dataset was utilized which contains 917 benchmarked pap smear cells of cervical with 26 attributes and two target variables for training and testing phase. We have adopted combination of convolutional network with variational autoencoder for data classification. The usage of variational autoencoder reduces the dimensionality of data for further processing with involvement of softmax layer for training. The results have been obtained over 917 cancerous image type pap smear cells, where 70% (642) allocated for training and remaining 30% (275) considered for test data set. The proposed architecture achieved variational accuracy of 99.2% with 2*2 filter size and 99.4% with 3*3 filter size using different epochs. The proposed hybrid variational convolutional autoencoder approach applied first time for cervical cancer diagnosis and performed better than traditional machine learning methods.
Circular RNAs (circRNAs) are a special kind of non-coding RNA. They play important regulatory role in diseases through interactions of miRNAs associated with the diseases. Due to their insensitivity to nucleases, they...
详细信息
Circular RNAs (circRNAs) are a special kind of non-coding RNA. They play important regulatory role in diseases through interactions of miRNAs associated with the diseases. Due to their insensitivity to nucleases, they are more stable than linear RNAs. It is thus imperative to integrate available information for predicting circRNA-disease associations in humans. Here, we propose a computational model to predict circRNA-disease associations based on accelerated attributed network embedding (AANE) algorithm and autoencoder(AE). First, we use AANE algorithm to extract low-dimensional features of circRNAs and diseases and then stacked autoencoder (SAE) to automatically extract in-depth features. The features obtained by AANE and the SAE are integrated and XGBoost is used as a binary classifier to get the predicted results. The proposed model has an average area under the receiver operating characteristic curve value of 0.8800 in 5-fold cross validation and 0.8988 in 10-fold cross validation. The factors that can affect the performance of the model are discussed and some common diseases are used as case studies. Results indicated that the model has great performance in predicting circRNA-disease associations. (c) 2021 Elsevier Inc. All rights reserved.
Heterogeneous cross-project defect prediction (HCPDP) is aimed at building a defect prediction model for the target project by reusing datasets from source projects, where the source project datasets and target projec...
详细信息
Heterogeneous cross-project defect prediction (HCPDP) is aimed at building a defect prediction model for the target project by reusing datasets from source projects, where the source project datasets and target project dataset have different features. Most existing HCPDP methods only remove redundant or unrelated features without exploring the underlying features of cross-project datasets. Additionally, when the transfer learning method is used in HCPDP, these methods ignore the negative effect of transfer learning. In this paper, we propose a novel HCPDP method called multi-source heterogeneous cross-project defect prediction (MHCPDP). To reduce the gap between the target datasets and the source datasets, MHCPDP uses the autoencoder to extract the intermediate features from the original datasets instead of simply removing redundant and unrelated features and adopts a modified autoencoder algorithm to make instance selection for eliminating irrelevant instances from the source domain datasets. Furthermore, by incorporating multiple source projects to increase the number of source datasets, MHCPDP develops a multi-source transfer learning algorithm to reduce the impact of negative transfers and upgrade the performance of the classifier. We comprehensively evaluate MHCPDP on five open source datasets;our experimental results show that MHCPDP not only has significant improvement in two performance metrics but also overcomes the shortcomings of the conventional HCPDP methods.
In this letter, we demonstrated the possibility of predicting full transistor current-voltage (IV) and capacitance-voltage (CV) curves using machines trained by Technology Computer-Aided Design (TCAD) generated data. ...
详细信息
In this letter, we demonstrated the possibility of predicting full transistor current-voltage (IV) and capacitance-voltage (CV) curves using machines trained by Technology Computer-Aided Design (TCAD) generated data. 3D FinFET IDVG and CGVG predictions are used as examples. The machine is constructed through manifold learning using autoencoder (AE) to extract the latent variables which are then correlated to physical parameters through 3rd-order polynomial regression. No device physics domain expertise is required in the machine learning process because there is no need to extract device metrics such as transconductance (g(m)) or Drain-Induced-BarrierLowering (DIBL) from the TCAD training data. We show that the machine can predict not just the full IV/ CV curves but also g(m) (1st derivative quantity) and DIBL (extracted from two machines trained by different data). Good results can be obtained even with < 50 training data. Our work shows that manifold learning is possible in device IV and CV to capture the complex physics and, thus, it is expected that it is possible to predict the IV/ CV of novel devices using limited experimental data before the underlying physics is well-understood.
autoencoder based methods are the majority of deep unsupervised outlier detection methods. However, these methods perform not well on complex image datasets and suffer from the noise introduced by outliers, especially...
详细信息
autoencoder based methods are the majority of deep unsupervised outlier detection methods. However, these methods perform not well on complex image datasets and suffer from the noise introduced by outliers, especially when the outlier ratio is high. In this paper, we propose a framework named Transformation Invariant autoencoder (TIAE), which can achieve stable and high performance on unsupervised outlier detection. First, instead of using a conventional autoencoder, we propose a transformation invariant autoencoder to do better representation learning for complex image datasets. Next, to mitigate the negative effect of noise introduced by outliers and stabilize the network training, we select the most confident inliers likely examples in each epoch as the training set by incorporating adaptive self-paced learning in our TIAE framework. Extensive evaluations show that TIAE significantly advances unsupervised outlier detection performance by up to 10% AUROC against other autoencoder based methods on five image datasets.
暂无评论