Enhancing Acoustic Scene Classification with Channel Attention Based on Deep Residual Learning Network

Deep residual networks (ResNets) can learn deep feature representation from hyperspectral images (HSIs), and therefore have been widely used for HSI classification. Despite their high accuracies, there still exist a lot of challenging cases, such as open world recognition, limited-sample learning and visualization of learned classification features, which cannot be well addressed. Most of the challenges in HSI classification can be attributed to the dependence on softmax based loss function and classifiers, which cause the lack of robustness for deep learning models and the hardness to visualize the learned classification features. To improve the robustness and achieve the visualization of learned classification features, we propose a novel learning framework called Residual Prototype Learning Network, a combination of residual network and prototypes learning mechanism. Under the framework, a prototype learning based loss function is proposed to enhance intra-class compactness and the inter-class separation of these feature representations; in addition, a prototype learning based classifier is simultaneously proposed to achieve the 2D or 3D visualization of the classification features. The effectiveness of our proposed learning framework is evaluated on several publicly available HSI benchmarks, and the experimental results show that our approach achieve better results than traditional softmax based ResNets.

Data-driven rolling-bearing fault diagnosis methods are mostly based on deep-learning models, and their multilayer nonlinear mapping capability can improve the accuracy of intelligent fault diagnosis. However, problems such as gradient disappearance occur as the number of network layers increases. Moreover, directly taking the raw vibration signals of rolling bearings as the network input results in incomplete feature extraction. In order to efficiently represent the state characteristics of vibration signals in image form and improve the feature learning capability of the network, this paper proposes fault diagnosis model MTF-ResNet based on a Markov transition field and deep residual network. First, the data of raw vibration signals are augmented by using a sliding window. Then, vibration signal samples are converted into two-dimensional images by MTF, which retains the time dependence and frequency structure of time-series signals, and a deep residual neural network is established to perform feature extraction, and identify the severity and location of the bearing faults through image classification. Lastly, experiments were conducted on a bearing dataset to verify the effectiveness and superiority of the MTF-ResNet model. Features learned by the model are visualized by t-SNE, and experimental results indicate that MTF-ResNet showed better average accuracy compared with several widely used diagnostic methods.

Nowadays, artificial intelligence has become more and more popular in daily life. In this paper, a ten-layer network model which is based on residual thought is proposed. And the deep learning framework Caffe is used to verify the proposed algorithm where a face dataset and a car dataset collected from multiple perspectives are employed. Results show that the proposed residual network model contributes to improve the accuracy of recognition.

Knowledge of molecular subtypes of gliomas can provide valuable information\nfor tailored therapies. This study aimed to investigate the use of deep\nconvolutional neural networks (DCNNs) for noninvasive glioma subtyping with\nradiological imaging data according to the new taxonomy announced by the World\nHealth Organization in 2016. Methods: A DCNN model was developed for the\nprediction of the five glioma subtypes based on a hierarchical classification\nparadigm. This model used three parallel, weight-sharing, deep residual\nlearning networks to process 2.5-dimensional input of trimodal MRI data,\nincluding T1-weighted, T1-weighted with contrast enhancement, and T2-weighted\nimages. A data set comprising 1,016 real patients was collected for evaluation\nof the developed DCNN model. The predictive performance was evaluated via the\narea under the curve (AUC) from the receiver operating characteristic analysis.\nFor comparison, the performance of a radiomics-based approach was also\nevaluated. Results: The AUCs of the DCNN model for the four classification\ntasks in the hierarchical classification paradigm were 0.89, 0.89, 0.85, and\n0.66, respectively, as compared to 0.85, 0.75, 0.67, and 0.59 of the radiomics\napproach. Conclusion: The results showed that the developed DCNN model can\npredict glioma subtypes with promising performance, given sufficient,\nnon-ill-balanced training data.\n

Abnormal target detection in hyperspectral remote sensing image is one of the hotspots in image research. The image noise generated in the detection process will lead to the decline of the quality of hyperspectral remote sensing image. In view of this, this paper proposes an abnormal target detection method of hyperspectral remote sensing image based on the convolution neural network. Firstly, the deep residual learning network model has been used to remove the noise in hyperspectral remote sensing image. Secondly, the spatial and spectral features of hyperspectral remote sensing images were used to optimize the clustering dictionary, and then the image segmentation containing target information is completed. Finally, the image was input into the deep convolution neural network with a dual classifier, and the network detects the abnormal target in the image. The test results of this algorithm show that the structural similarity of the denoised image is higher than 0.86, which shows that this method has good noise reduction performance, image details will not damage, segmentation effect is good, and it can obtain high-definition target image information and accurately detect abnormal targets in the image.

Reliable automatic target segmentation in Synthetic Aperture Radar (SAR) imagery has played an important role in the SAR fields. Different from the traditional methods, Spectral Residual (SR) and CFAR detector, with the recent advancements in machine learning theory, new methods based on machine learning show better performance in SAR target segmentation. In this paper, we proposed a deep deformable residual learning network for target segmentation, which attempts to preserve the precise contour of the target. The deformable convolutional layers and residual learning block are applied in this technique, which could extract and preserve the geometric information of the target as much as possible. The proposed method is evaluated based on the Moving and Stationary Target Acquisition and Recognition (MSTAR) dataset, and the experimental results have shown the superiority of the proposed network for the precise segmentation of the target.

Due to the favorable denoising performance of the discriminative learning model, a large number of discriminative learning models have been designed to remove noise. However, most discriminative learning models can only deal with the noise that already exists in the training data. In this paper, a multi-scale trainable deep residual convolutional neural network (DCMSNet) based on dilated convolution is proposed. DCMSNet consists of a chain of dilated convolution layers, convolution layers, normalized multi-scale convolution blocks (BNMCBlock), multi-scale convolution blocks (MCBlock) and multi-scale dropout convolution blocks (MCDBlock). The use of dilated convolution can avoid over-parameterization problems caused by too deep networks. In order to capture more feature information during feature extraction and image reconstruction, we designed novel BNMCBlock and MCBlock. In order to reduce the degree of coupling between image features and improve the generalization ability of the network, we also designed a MCDBlock. Meanwhile, residual learning, batch normalization and dropout are utilized to speed up the training process and boost the denoising performance. Unlike existing denoising models, DCMSNet is able to remove different degrees of noise. Compared with state-of-the-art image denoising methods, DCMSNet has achieved relatively competitive results.

As an important part of the power system, gas insulated switchgear (GIS) will cause serious failures once they break down, threatening the safety of the entire power grid. In the construction of the Ubiquitous Power Internet of Things (UPIoT), the intelligent terminal taking online monitoring as the means keeps the equipment fault samples and forms the sample database, which is of great significance to discover the latent insulation defects of GIS and take necessary measures in advance to ensure the safe and reliable operation of power grid. Aiming at the sample database provided by the intelligent terminal of the Internet of things, this paper proposes a method of GIS partial discharge (PD) using the depth residual network, which effectively improves the accuracy of model recognition. Although the comprehensiveness of the sample has been solved, as a transitional stage, the sample size is relatively small. Therefore, this paper uses transfer learning to solve the problem of high accuracy under the sample. In order to compare the state of art performance of the proposed method, some traditional convolutional networks such as LeNet, AlexNet, and VGG16 are used for comparison. After verification, the recognition accuracy of the deep residual network proposed in this paper is 94.6%, which is significantly higher than other models. At the same time, the parameter amount and storage space of the deep residual network are also significantly lower than those of other networks, further verifying that the model has a broad application space in the UPIoT context.

Various microfossils from the early Cambrian provide crucial clues for understanding the Cambrian explosion and the origin of animal phyla. However, specimens with important anatomical structures are extremely rare and the efficiency of retrieving such fossils by traditional manual selection under a microscope is quite low. Such a contradiction has hindered breakthroughs in micropaleontology for a long time. Here, we propose a solution for identifying specific taxa of Cambrian microfossils using only a few available specimens by transferring a model pre-trained on natural image datasets to the field of paleontological artificial intelligence. The method employs a 34-layer deep residual neural network as the underlying framework, migrates the ImageNet pre-trained model, freezes the low-layer network parameters and retrains the high-layer parameters to build a microfossil image recognition model. We built training sets with randomly selected images of varied number for each taxon. Our experiments show that the average recognition accuracy for specific taxa of Cambrian microfossils (50 images for each taxon) is higher than 0.97 and it can reach 0.85 with only three training samples per taxon. Comparative analyses indicate that our results are much better than those of various prevalent methods, such as the transpose convolutional neural network (TCNN). This demonstrates the feasibility of using natural images (ImageNet) for the training of microfossil recognition models and provides a promising tool for the discovery of rare fossils.

Aiming at the problem that the difference in fault data distribution caused by the variation of gear and bearing conditions during the operation of rotating machinery affects the accuracy of fault diagnosis, a multi-condition fault diagnosis approach based on whale optimization algorithm (WOA), variational mode decomposition (VMD), and deep residual shrinkage network (DRSN) is proposed. Firstly, WOA is employed to optimize the parameters of the decomposition level and the penalty factor value in VMD, enhancing its noise resistance. Subsequently, VMD processing yields several modal components, each characterized by a permutation entropy feature vector, ensuring optimal feature extraction. DRSN is then utilized, employing an attention mechanism to adaptively set soft threshold functions, thereby improving gear fault diagnosis accuracy under multiple operating conditions. Finally, the proposed method’s performance was validated through the collection of gear running signals under various conditions, and the fault diagnosis accuracy was used as the model evaluation index to conduct comparative experiments and ablation experiments. The experimental results indicate that the WOA-VMD feature extraction and DRSN fault diagnosis techniques outperform traditional methods in multi-condition gear fault diagnosis. This study provides a new effective approach for multi-condition diagnosis of rotating machinery.

Subject to intricate lighting conditions and environmental nuances, including suspended particles, underwater images often grapple with color distortion, background blurriness, and limited resolution. To cope with these challenges and elevate the quality of underwater image reconstruction and restoration, we introduce a super-resolution reconstruction algorithm for underwater images, leveraging a Very Deep Super-Resolution (VDSR) network enrich ed with deep residual learning techniques, namely, Very Deep Residual SuperResolution Network (VDRSRN). This methodology endeavors to meticulously fit the intricate features of the underwater milieu, forge deeper mapping associations, and attain superior model generalization. By integrating residual block structures in tandem with an alternating connection strategy, we amplify the network's depth, thereby augmenting its proficiency in portraying intricate underwater scenes. Additionally, this approach effectively counters the pervasive issues of gradient vanishing and model degradation inherent in deep networks, facilitated by residual learning mechanisms. The alternating configuration of identity residual blocks and convolutional residual blocks retains the stability benefits of identity connections while fostering efficient feature reuse and transmission, significantly mitigating the loss of low-level features within deep networks. Experimental results demonstrate that the proposed algorithm outperforms comparative algorithms in both subjective and objective evaluations, with substantial preservation of texture, edges, and other details in reconstructed underwater images.

Cascaded deep residual learning network for single image dehazing

Some industrial processes take place in confined settings only observable by sensors, e.g. infrared (IR) cameras. Drying processes take place while a material is transported by means of a conveyor through a black box equipped with internal IR cameras. While such sensors deliver data at high rates, this is beyond what human operators can analyze and calls for automation. Inspired by numerous implementations monitoring techniques that analyse IR images using deep learning, this paper shows how they can be applied to the confined microwave drying of porous foams, with benchmarking their effectiveness at condition monitoring to conduct fault detection. Convolutional neural networks, derived transfer learning, and deep residual neural network methods are already regarded as cutting-edge and are studied here, using a set of conventional approaches for comparative evaluation. Our comparison shows that state-of-the-art deep learning techniques significantly benefit condition monitoring, providing an increase in fault finding accuracy of up to 48% over conventional methods. Nevertheless, we also found that derived transfer learning and deep residual network techniques do not in our case yield increased performance over normal convolutional neural networks.

Deep learning (also known as deep structured learning, hierarchical learning, or deep machine learning) is a branch of machine learning based on a set of algorithms that attempt to model high-level abstractions in data by using a deep graph with multiple processing layers, composed of multiple linear and nonlinear transformations. Deep learning has been characterized as a class of machine learning algorithms with the following characteristics [257]: • They use a cascade of many layers of nonlinear processing units for feature extraction and transformation. Each successive layer uses the output from the previous layer as input. The algorithms may be supervised or unsupervised and applications include pattern analysis (unsupervised) and classification (supervised). • They are based on the (unsupervised) learning of multiple levels of features or representations of the data. Higher-level features are derived from lower-level features to form a hierarchical representation. • They are part of the broader machine learning field of learning representations of data. • They learn multiple levels of representations that correspond to different levels of abstraction; the levels form a hierarchy of concepts. These definitions have in common: multiple layers of nonlinear processing units and the supervised or unsupervised learning of feature representations in each layer, with the layers forming a hierarchy from low-level to high-level features. Various deep learning architectures such as deep neural networks, convolutional deep neural networks, deep belief networks (DBN), and recurrent neural networks have been applied to fields like computer vision, automatic speech recognition, natural language processing, audio recognition, and bioinformatics where they have been shown to produce state-of-the-art results on various tasks. In this chapter we start in Section 7.1 with an introduction, giving a brief history of this field, the relevant literature, and its applications. Then we study some basic concepts of deep learning such as convolutional neural networks, recurrent neural networks, backpropagation algorithm, restricted Boltzmann machines, and deep learning networks in Section 7.2. Then we illustrate three examples for Apache Spark implementation for mobile big data (MBD), user moving pattern extraction, and combination with nonparametric Bayesian learning, respectively, in Sections 7.3 through 7.5. Finally, we have summary in Section 7.6.

Two key challenges lie in the facial attractiveness computation research: the lack of discriminative face representations, and the scarcity of sufficient and complete training data. Motivated by recent promising work in face recognition using deep neural networks to learn effective features, the first challenge is expected to be addressed from a deep learning point of view. A very deep residual network is utilized to enable automatic learning of hierarchical aesthetics representation. The inspiration to deal with the second challenge comes from the natural representation of the training data, where each training face can be associated with a label (score) distribution given by human raters rather than a single label (average score). This paper, therefore, recasts facial attractiveness computation as a label distribution learning problem. Integrating these two ideas, an end-to-end attractiveness learning framework is established. We also perform feature-level fusion by incorporating the low-level geometric features to further improve the computational performance. Extensive experiments are conducted on a standard benchmark, the SCUT-FBP dataset, where our approach shows significant advantages over the other state-of-the-art work.