Unfertilized duck eggs not removed prior to incubation will deteriorate quickly, posing a risk of contaminating the normally fertilized duck eggs. Thus, detecting the fertilization status of breeding duck eggs as earl...
详细信息
Unfertilized duck eggs not removed prior to incubation will deteriorate quickly, posing a risk of contaminating the normally fertilized duck eggs. Thus, detecting the fertilization status of breeding duck eggs as early as possible is a meaningful and challenging task. Most existing work usually focus on the characteristics of chicken eggs during mid-term hatching. However, little attention has been paid to the detection for duck eggs prior to incubation. In this paper, we present a novel hybrid deep learning detection framework for the fertilization status of pre -incubation duck eggs, termed CVAE-DF, based on visible/near-infrared (VIS/NIR) transmittance spectroscopy. The framework comprises the encoder of a convolutional variational autoencoder (CVAE) and an improved deep forest (DF) model. More specifically, we first collected transmittance spectral data (400 -1000 nm) of 255 duck eggs before hatching. The multiplicative scatter correction (MSC) method was then used to eliminate noise and extraneous information of the raw spectral data. Two efficient data augmentation methods were adopted to provide sufficient data. After that, CVAE was applied to extract representative features and reduce the feature dimension for the detection task. Finally, an improved DF model was employed to build the classification model on the enhanced feature set. The CVAE-DF model achieved an overall accuracy of 95.94 % on the test dataset. These experimental results in terms of four metrics demonstrate that our CVAE-DF method outperforms the traditional methods by a significant margin. Furthermore, the results also indicate that CVAE holds great promise as a novel feature extraction method for the VIS/NIR spectral analysis of other agricultural products. It is extremely beneficial to practical engineering.
In recent years, researchers have extensively explored the application of drive -by inspection technology for bridge damage assessment. This approach involves using the response of a sensing vehicle to identify damage...
详细信息
In recent years, researchers have extensively explored the application of drive -by inspection technology for bridge damage assessment. This approach involves using the response of a sensing vehicle to identify damage. However, many existing methods rely on data collected from both healthy and damaged bridge conditions, which may not always be available. Therefore, this study introduces a fully unsupervised computer vision -based methodology for bridge structural health monitoring (BSHM) using drive -by inspection. It analyzes the time-frequency domain of a two -axle vehicle's response by deriving a novel formulation for the contact point response from vehicle axles. The axles signals are then processed through subtraction, filtering, and decomposition using empirical Fourier decomposition with an improved segmentation approach based on the Savitzky-Golay filter (SGEFD). Relevant Intrinsic Mode Functions (IMF) are extracted as features representing damage, and the Wavelet Synchro-squeezed transform (WSST) is obtained from these features and used as input for the damage assessment algorithm. The performance of two state-of-the-art unsupervised generative machine learning methods, namely convolutional variational autoencoders (CVAE) and convolutional adversarial autoencoders (CAAE), is compared for the damage assessment task. These methods are trained solely with the residual WSST obtained from the vehicle responses when traversing a bridge in its reference state. A damage index (DI) is defined based on the measured error between the original and reconstructed images, and a damage threshold is calculated from the DI distribution of samples from the benchmark bridge state. During testing, the error between the original and reconstructed WSST is compared to the damage threshold, enabling the classification of new samples as healthy or damaged. The methodology is evaluated using both numerical and experimental vehicle-bridge interaction (VBI) models, considering various
Characterization of fracture network is essential for understanding groundwater flow and solute transport, as well as waste storage. Deep -learning based ensemble smoother methods have proven to be effective in estima...
详细信息
Characterization of fracture network is essential for understanding groundwater flow and solute transport, as well as waste storage. Deep -learning based ensemble smoother methods have proven to be effective in estimating hydraulic parameters in porous media. Compared to porous media, fracture fields are highly heterogeneous and typically non -Gaussian distributed, making the estimation of the fracture field from sparse borehole data extremely difficult. In this paper, we developed a joint hydrogeophysical inversion framework to improve the characterization of fracture networks. We first trained a convolutional variational autoencoder (CVAE) network to parameterize the fracture field, and then integrated with the ensemble smoother with multiple data assimilation (ESMDA) method to infer the fracture distribution by incorporating multiple datasets. Two numerical cases with different complexity were considered to assess the ability of the proposed joint inversion framework. The results show that the proposed hydrogeophyscial inversion framework can capture the main features of the fracture field. By integrating both the pressure and SP data, the fracture field can be reconstructed with an improved accuracy and reduced estimation uncertainty.
Visual defect inspection and classification are significant steps of most manufacturing processes in the semiconductor and electronics industries. Known and unknown defects on wafer maps tend to cluster, and these spa...
详细信息
Visual defect inspection and classification are significant steps of most manufacturing processes in the semiconductor and electronics industries. Known and unknown defects on wafer maps tend to cluster, and these spatial patterns provide valuable process information for supporting manufacturing in determining the root causes of abnormal processes. In previous studies, data augmentation-based deep learning (DL) techniques were most commonly used for the identification of wafer map defect patterns (WMDP). Data augmentation is an effective technique for improving the accuracy of modern image classifiers. However, current data augmentation implementations were manually designed for the WMDP problem. In this study, we propose a DL-based method with automatic data augmentation for the WMDP task. Basically, it focuses on learning effective discriminative features, from wafer maps, through a deep network structure. The network consists of a convolution-based variationalautoencoder (CVAE) sequentially. First, we pre-trained the CVAE on large training data in an unsupervised manner. Second, we fine-tuned the encoder of the CVAE, which was followed by a neural network (NN) classifier, in a supervised manner. Additionally, we describe a simple procedure for automatically searching for improved data augmentation policies. The policy mainly consists of five image processing functions: rotation, flipping, shifting, shearing range, and zooming. The effectiveness of the proposed method was demonstrated through experimental results obtained from a simulation dataset and a real-world wafer map dataset (WM-811K). This study provides guidance for the application of deep learning in semiconductor manufacturing processes to improve product quality and yield.
More and more scientific disciplines are using deep techniques for the automatic classification of massive high dimensionality unlabeled data. Among these disciplines, the classification of partial discharge (PD) patt...
详细信息
ISBN:
(纸本)9781665419703
More and more scientific disciplines are using deep techniques for the automatic classification of massive high dimensionality unlabeled data. Among these disciplines, the classification of partial discharge (PD) patterns is one that represents major challenges in the field of hydrogenerator diagnosis. This paper proposes a method of comparison of five classification topology based on a single convolutional variational autoencoder (CVAE) and ten classifiers. The comparison is based on five cases exploiting all the same database, but using five different feature extraction rules to create the input vectors of the neural networks. These feature extraction rules are based on the expert judgement and are automatically computed in the preprocessing stage. Analysis of the output of all classifiers for each topology suggests that the accuracy level of the classification can be significantly improved by refining the feature extraction rules. Moreover, the visualization of the 2D latent space from the CVAE also suggests that the accuracy level can be even further improved if the whole dataset is considered instead of a smaller reference dataset randomly selected. Results raise many questions about the performance of feature extraction rules and the possibilities to better handling classification of large databases such as the one of PD measurement files used for hydrogenerator diagnosis.
Several problems can be encountered in the design of autonomous vehicles. Their software is organized into three main layers: perception, planning, and actuation. The planning layer deals with the sort and long-term s...
详细信息
Several problems can be encountered in the design of autonomous vehicles. Their software is organized into three main layers: perception, planning, and actuation. The planning layer deals with the sort and long-term situation prediction, which are crucial for intelligent vehicles. Whatever method is used to make forecasts, vehicles' dynamic environment must be processed for accurate long-term forecasting. In the present article, a method is proposed to preprocess the dynamic environment in a freeway traffic situation. The method uses the structured data of surrounding vehicles and transforms it to an occupancy grid which a convolutional variational autoencoder (CVAE) processes. The grids (2048 pixels) are compressed to a 64-dimensional latent vector by the encoder and reconstructed by the decoder. The output pixel intensities are interpreted as probabilities of the corresponding field is occupied by a vehicle. This method's benefit is to preprocess the structured data of the dynamic environment and represent it in a lower-dimensional vector that can be used in any further tasks built on it. This representation is not handmade or heuristic but extracted from the database patterns in an unsupervised way.
One of the major differences between medical doctor training and machine learning is that doctors are trained to recognize normal/healthy anatomy first. Knowing the healthy appearance of anatomy structures helps docto...
详细信息
ISBN:
(纸本)9783030322267;9783030322250
One of the major differences between medical doctor training and machine learning is that doctors are trained to recognize normal/healthy anatomy first. Knowing the healthy appearance of anatomy structures helps doctors to make better judgement when some abnormality shows up in an image. In this study, we propose a normal appearance autoencoder (NAA), that removes abnormalities from a diseased image. This autoencoder is semi-automatically trained using another partial convolutional in-paint network that is trained using healthy subjects only. The output of the autoencoder is then fed to a segmentation net in addition to the original input image, i.e. the latter gets both the diseased image and a simulated healthy image where the lesion is artificially removed. By getting access to knowledge of how the abnormal region is supposed to look, we hypothesized that the segmentation network could perform better than just being shown the original slice. We tested the proposed network on the LIDC-IDRI dataset for lung cancer detection and segmentation. The preliminary results show the NAA approach improved segmentation accuracy substantially in comparison with the conventional U-Net architecture.
We present a novel approach to derive robust speech representations for automatic speech recognition (ASR) systems. The proposed method uses an unsupervised data-driven modulation filter learning approach that preserv...
详细信息
ISBN:
(纸本)9781479981311
We present a novel approach to derive robust speech representations for automatic speech recognition (ASR) systems. The proposed method uses an unsupervised data-driven modulation filter learning approach that preserves the key modulations of speech signal in spectro-temporal domain. This is achieved by a deep generative modeling framework to learn modulation filters using convolutional variational autoencoder (CVAE). A skip connection based CVAE enables the learning of multiple irredundant modulation filters in the time and frequency modulation domain using temporal and spectral trajectories of input spectrograms. The learnt filters are used to process the spectrogram features for ASR training The ASR experiments are performed on Aurora-4 (additive noise with channel artifact) and CHiME-3 (additive noise with reverberation) databases. The results show significant improvements for the proposed CVAE model over the baseline features as well as other robust front-ends (average relative improvements of 9% in word error rate over base-line features on Aurora-4 database and 23% on CHiME-3 database). In addition, the performance of the proposed features is highly beneficial for semi-supervised training of ASR when reduced amounts of labeled training data are available (average relative improvements of 29% over baseline features on Aurora-4 database with 30% of the labeled training data).
In this paper, we propose a deep representation learning approach using the raw speech waveform in an unsupervised learning paradigm. The first layer of the proposed deep model performs acoustic filtering while the su...
详细信息
In this paper, we propose a deep representation learning approach using the raw speech waveform in an unsupervised learning paradigm. The first layer of the proposed deep model performs acoustic filtering while the subsequent layer performs modulation filtering. The acoustic filterbank is implemented using cosine-modulated Gaussian filters whose parameters are learned. The modulation filtering is performed on log transformed outputs of the first layer and this is achieved using a skip connection based architecture. The outputs from this two layer filtering are fed to the variationalautoencoder model. All the model parameters including the filtering layers are learned using the VAE cost function. We employ the learned representations (second layer outputs) in a speech recognition task. Experiments are conducted on Aurora-4 (additive noise with channel artifact) and CHiME-3 (additive noise with reverberation) databases. In these experiments, the learned representations from the proposed framework provide significant improvements in ASR results over the baseline filterbank features and other robust front-ends (average relative improvements of 16% and 6% in word error rate over baseline features on clean and multi-condition training, respectively on Aurora-4 dataset, and 21% over the baseline features on CHiME-3 database).
暂无评论