High Discriminative Features Research Articles

A novel graph-based multiple kernel learning framework for hyperspectral image classification

ABSTRACT Multiple kernel learning (MKL) is an efficient way to improve hyperspectral image classification with few training samples by integrating spectral and spatial features. Nonetheless, presenting a MKL algorithm to capture significant information and eliminate redundancy is indispensable. In this paper, a novel graph-based MKL method (GMKL) is introduced to enhance the performance of the MKL algorithm and achieve high discriminative features, resulting in maximum separability. To do so, an informative low-rank graph is determined via a new algorithm called block diagonal representation that takes into account both local and global data structures. Then, in lieu of a time-consuming search for finding the optimal combination of the basic kernels, unlike the prior works, an optimal projective direction is acquired using graph theory. Various basic kernels generated from superpixels and attribute profiles are projected to produce a discriminative combination of kernels. Eventually, the optimum kernel is given to the support vector machine classifier. Experiments have been conducted on two popular hyperspectral datasets (Indian Pines and Pavia University). For the datasets, the proposed algorithm gained 90.43% and 99.28% overall accuracy, respectively, using only 1% training samples, which is much better than the state-of-the-art MKL algorithm and deep learning-based approaches with lower computational time.

PFP-HOG: Pyramid and Fixed-Size Patch-Based HOG Technique for Automated Brain Abnormality Classification with MRI.

Detecting neurological abnormalities such as brain tumors and Alzheimer's disease (AD) using magnetic resonance imaging (MRI) images is an important research topic in the literature. Numerous machine learning models have been used to detect brain abnormalities accurately. This study addresses the problem of detecting neurological abnormalities in MRI. The motivation behind this problem lies in the need for accurate and efficient methods to assist neurologists in the diagnosis of these disorders. In addition, many deep learning techniques have been applied to MRI to develop accurate brain abnormality detection models, but these networks have high time complexity. Hence, a novel hand-modeled feature-based learning network is presented to reduce the time complexity and obtain high classification performance. The model proposed in this work uses a new feature generation architecture named pyramid and fixed-size patch (PFP). The main aim of the proposed PFP structure is to attain high classification performance using essential feature extractors with both multilevel and local features. Furthermore, the PFP feature extractor generates low- and high-level features using a handcrafted extractor. To obtain the high discriminative feature extraction ability of the PFP, we have used histogram-oriented gradients (HOG); hence, it is named PFP-HOG. Furthermore, the iterative Chi2 (IChi2) is utilized to choose the clinically significant features. Finally, the k-nearest neighbors (kNN) with tenfold cross-validation is used for automated classification. Four MRI neurological databases (AD dataset, brain tumor dataset 1, brain tumor dataset 2, and merged dataset) have been utilized to develop our model. PFP-HOG and IChi2-based models attained 100%, 94.98%, 98.19%, and 97.80% using the AD dataset, brain tumor dataset1, brain tumor dataset 2, and merged brain MRI dataset, respectively. These findings not only provide an accurate and robust classification of various neurological disorders using MRI but also hold the potential to assist neurologists in validating manual MRI brain abnormality screening.

A Hybrid Approach Based on GAN and CNN-LSTM for Aerial Activity Recognition

Unmanned aerial vehicles (UAVs), known as drones, have played a significant role in recent years in creating resilient smart cities. UAVs can be used for a wide range of applications, including emergency response, civil protection, search and rescue, and surveillance, thanks to their high mobility and reasonable price. Automatic recognition of human activity in aerial videos captured by drones is critical for various tasks for these applications. However, this is difficult due to many factors specific to aerial views, including camera motion, vibration, low resolution, background clutter, lighting conditions, and variations in view. Although deep learning approaches have demonstrated their effectiveness in a variety of challenging vision tasks, they require either a large number of labelled aerial videos for training or a dataset with balanced classes, both of which can be difficult to obtain. To address these challenges, a hybrid data augmentation method is proposed which combines data transformation with the Wasserstein Generative Adversarial Network (GAN)-based feature augmentation method. In particular, we apply the basic transformation methods to increase the amount of video in the database. A Convolutional Neural Network–Long Short-Term Memory (CNN-LSTM) model is used to learn the spatio-temporal dynamics of actions, then a GAN-based technique is applied to generate synthetic CNN-LSTM features conditioned on action classes which provide a high discriminative spatio-temporal features. We tested our model on the YouTube aerial database, demonstrating encouraging results that surpass those of previous state-of-the-art works, including an accuracy rate of 97.83%.

P2FEViT: Plug-and-Play CNN Feature Embedded Hybrid Vision Transformer for Remote Sensing Image Classification

Remote sensing image classification (RSIC) is a classical and fundamental task in the intelligent interpretation of remote sensing imagery, which can provide unique labeling information for each acquired remote sensing image. Thanks to the potent global context information extraction ability of the multi-head self-attention (MSA) mechanism, visual transformer (ViT)-based architectures have shown excellent capability in natural scene image classification. However, in order to achieve powerful RSIC performance, it is insufficient to capture global spatial information alone. Specifically, for fine-grained target recognition tasks with high inter-class similarity, discriminative and effective local feature representations are key to correct classification. In addition, due to the lack of inductive biases, the powerful global spatial context representation capability of ViT requires lengthy training procedures and large-scale pre-training data volume. To solve the above problems, a hybrid architecture of convolution neural network (CNN) and ViT is proposed to improve the RSIC ability, called P2FEViT, which integrates plug-and-play CNN features with ViT. In this paper, the feature representation capabilities of CNN and ViT applying for RSIC are first analyzed. Second, aiming to integrate the advantages of CNN and ViT, a novel approach embedding CNN features into the ViT architecture is proposed, which can make the model synchronously capture and fuse global context and local multimodal information to further improve the classification capability of ViT. Third, based on the hybrid structure, only a simple cross-entropy loss is employed for model training. The model can also have rapid and comfortable convergence with relatively less training data than the original ViT. Finally, extensive experiments are conducted on the public and challenging remote sensing scene classification dataset of NWPU-RESISC45 (NWPU-R45) and the self-built fine-grained target classification dataset called BIT-AFGR50. The experimental results demonstrate that the proposed P2FEViT can effectively improve the feature description capability and obtain outstanding image classification performance, while significantly reducing the high dependence of ViT on large-scale pre-training data volume and accelerating the convergence speed. The code and self-built dataset will be released at our webpages.

Deep Residual SVM: A Hybrid Learning Approach to obtain High Discriminative Feature for Land Use and Land Cover Classification

Classification of multispectral images is impacted by challenges such as inadequate training samples, limited ground truth, and complex spatiotemporal dependencies. The accuracy of classifiers due to the lack of training samples is also compounded by the class imbalance problem in spatial analytics.This paper proposes a novel idea of exploiting the deep residual networks trained on millions of images for high discriminative feature representations. These derived representations are classified using bias-corrected Adam optimizer-based Support Vector Machine classifiers. The intuition behind this approach is the significance of representations in the last layers of the pre-trained network that contributes to the classification accuracy. These encapsulated representations from the pre-trained network can outperform traditional feature representations. Residual networks solved the vanishing gradients problem in deep learning and improved classification accuracy. Hence, this work explores the combined potential of the representation power of deep residual pre-trained networks and the classification ability of Support Vector Machines. The model implementation experiments have been conducted using two publicly available benchmark datasets, Eurosat and UC Merced for Land Use and Land Cover Classification. The proposed model DRSVM (Deep Residual Support Vector Machines) demonstrated higher efficiency in terms of computational parameters and time complexity than conventional convolutional neural networks, pre-trained networks, and Support Vector Machines with comparable accuracy. Data augmentation techniques are used further for enhancing the performance of the model. Conventional supervised deep learning algorithms like CNN overfit in dense layers in case of inadequate training data. This hybrid approach alleviates overfitting issues by eliminating dense layers and replacing the Softmax function with Support Vector Machines for classification.

Hyperspectral Target Detection via Spectral Aggregation and Separation Network With Target Band Random Mask

Hyperspectral target detection (HTD) is a pixel-wise detection method based on limited prior targets and spectral differences, which has been widely studied and applied in many fields. Recently, deep learning (DL) plays an important role in hyperspectral imagery (HSI) processing. However, for HTD, the severe lack of class-balanced training sets is an enormous challenge. Meanwhile, it is difficult to suppress backgrounds while highlighting targets through the deep network. To address these issues, we propose a spectral aggregation and separation network (SASN) with a target band random mask (TBRM) for HTD in this paper. For the training sets of SASN, a multifarious representative background selection strategy (MRBS) is first proposed to obtain a multifarious and representative background training set. Next, aiming at the notorious class imbalance, a data augmentation (DA) method, TBRM, is proposed to generate adequate target training set by repeating randomly zero-masking the spectral bands of a prior target. Subsequently, in the training of SASN, residual connection and squeeze-and-excitation (SE) channel attention mechanism are applied to fully extract high discriminative features and nonlinear ones in the spectra. Besides, to better separate the targets and backgrounds, a triplet-soft loss function is presented, which makes the training in the direction of spectral separation of background samples from both the prior target and target samples. During testing, the trained SASN distinguishes the spectral similarities and differences simultaneously for highlighting targets and suppressing backgrounds. Moreover, extensive experimental results validate that the proposed method has superior detection performances, background suppression capacity, and separability compared with ten cutting-edge HTD algorithms on six benchmark HSI datasets.

Approximate zero-crossing: a new interpretable, highly discriminative and low-complexity feature for EEG and iEEG seizure detection

Objective. Long-term monitoring of people with epilepsy based on electroencephalography (EEG) and intracranial EEG (iEEG) has the potential to deliver key clinical information for personalised epilepsy treatment. More specifically, in outpatient settings, the available solutions are not satisfactory either due to poor classification performance or high complexity to be executed in resource-constrained devices (e.g. wearable systems). Therefore, we hypothesize that obtaining high discriminative features is the main avenue to improve low-complexity seizure-detection algorithms. Approach. Inspired by how neurologists recognize ictal EEG data, and to tackle this problem by targeting resource-constrained wearable devices, we introduce a new interpretable and highly discriminative feature for EEG and iEEG, namely approximate zero-crossing (AZC). We obtain AZC by applying a polygonal approximation to mimic how our brain selects prominent patterns among noisy data and then using a zero-crossing count as a measure of the dominating frequency. By employing Kullback–Leiber divergence, leveraging CHB-MIT and SWEC-ETHZ iEEG datasets, we compare the AZC discriminative power against a set of 56 classical literature features (CLF). Moreover, we assess the performances of a low-complexity seizure detection method using only AZC features versus employing the CLF set. Main results. Three AZC features obtained with different approximation thresholds are among the five with the highest median discriminative power. Moreover, seizure classification based on only AZC features outperforms an equivalent CLF-based method. The former detects 102 and 194 seizures, against 99 and 161 for the latter (CHB-MIT and SWEC-ETHZ, respectively). Moreover, the AZC-based method keeps a similar false-alarm rate (i.e. an average of 2.1 and 1.0, against 2.0 and 0.5, per day). Significance. We propose a new feature and demonstrate its capability in seizure classification for both scalp and intracranial EEG. We envision the use of such a feature to improve outpatient monitoring with resource-constrained devices.

Reconstruct face from features based on genetic algorithm using GAN generator as a distribution constraint

Face recognition based on deep convolutional neural networks (CNN) shows superior accuracy performance attributed to the high discriminative features extracted. Yet, the security and privacy of the extracted features from deep learning models (deep features) have often been overlooked. This paper proposes the reconstruction of face images from deep features without accessing the CNN network configurations as a constrained optimization problem. Such optimization minimizes the distance between the features extracted from the original face image and the reconstructed face image. Instead of directly solving the optimization problem in the image space, we innovatively reformulate the problem by looking for a latent vector of a generative adversarial networks (GAN) generator, then use it to generate the face image. The GAN generator serves a dual role in this novel framework, i.e., face distribution constraint of the optimization goal and a face generator. To solve this optimization problem, We present an optimization approach based on a Genetic Algorithm. On top of the novel optimization task, we also propose an attack pipeline to impersonate the target user based on the generated face image. Our results show that the generated face images can achieve a state-of-the-art successful attack rate of 99.33% on Labeled Faces in the Wild (LFW) under type-I attack at a false accept rate of 0.1%. Our work sheds light on biometric deployment to meet privacy-preserving and security policies.

A hybrid-attention-ConvLSTM-based deep learning architecture to extract modal frequencies from limited data using transfer learning

This paper leverages each pixel of a picture acquired from a video camera, in which structural dynamic information is contained, in order to decompose spatiotemporal information from such a non-contact virtual sensor array in the same way as traditional accelerometers to extract structural modal frequencies. Attention-based deep neural network architecture is proposed in this work to better visualize the dynamic properties of structures in the existence of noise with a high resolution. The work combines CNNs and Recurrent Neural Networks (RNNs) to predict modal frequencies of structures from a series of consecutive images. High discriminative features of video frames are firstly extracted using the CNN, and then Conv-Long Short-Term Memory (ConvLSTM) is applied to further process the extracted features to capture the temporal dynamics in videos. The attention mechanisms are embedded in the network to ensure the model learns to focus selectively on those frames containing system dynamics. In particular, the proposed computer vision-based deep learning model takes the video of a vibrating structure as the input and successfully estimates the modal frequencies. Transfer learning is applied to cohere the knowledge learned from publicly available datasets to a much more sophisticated structure and estimate the resonant frequencies. The proposed algorithm optimizes the filter design for video processing in a fully automated way without any human intervention and can generalize and transfer that learned information to more complex structures. The model is trained using publicly available generic baseline data (Dataset A) consisting of several simple beam structures with different material properties and sizes and transferred the learned knowledge to unseen data (Dataset B) consisting of an independent turbine blade. It is concluded that the newly proposed method is more autonomous, accurate, and capable of generalizing the model to a new independent dataset using a transfer learning strategy, and the most advantage of the proposed approach is that the trained deep learning architecture has the capability of estimating the resonant frequencies for independent structures and extending the resonant frequency estimations to higher modes.

MDGF-MCEC: a multi-view dual attention embedding model with cooperative ensemble learning for CircRNA-disease association prediction.

Circular RNA (circRNA) is closely involved in physiological and pathological processes of many diseases. Discovering the associations between circRNAs and diseases is of great significance. Due to the high-cost to verify the circRNA-disease associations by wet-lab experiments, computational approaches for predicting the associations become a promising research direction. In this paper, we propose a method, MDGF-MCEC, based on multi-view dual attention graph convolution network (GCN) with cooperative ensemble learning to predict circRNA-disease associations. First, MDGF-MCEC constructs two disease relation graphs and two circRNA relation graphs based on different similarities. Then, the relation graphs are fed into a multi-view GCN for representation learning. In order to learn high discriminative features, a dual-attention mechanism is introduced to adjust the contribution weights, at both channel level and spatial level, of different features. Based on the learned embedding features of diseases and circRNAs, nine different feature combinations between diseases and circRNAs are treated as new multi-view data. Finally, we construct a multi-view cooperative ensemble classifier to predict the associations between circRNAs and diseases. Experiments conducted on the CircR2Disease database demonstrate that the proposed MDGF-MCEC model achieves a high area under curve of 0.9744 and outperforms the state-of-the-art methods. Promising results are also obtained from experiments on the circ2Disease and circRNADisease databases. Furthermore, the predicted associated circRNAs for hepatocellular carcinoma and gastric cancer are supported by the literature. The code and dataset of this study are available at https://github.com/ABard0/MDGF-MCEC.

Research on the shaft orbit of hydropower units based on convolutional neural network and the fine-grained indicators

Shaft orbit is one of the most representative characteristics of vibration signals which reflect the healthy and breakdown status of hydropower units. However, the traditional pattern recognition methods consisting of feature extraction and classifier could not been able to identify the kinds of faults and the degree of severity effectively. In this paper, a novel intelligent method based on convolutional neural network (CNN) and the fine-grained indicators is proposed for auto-diagnosis of faults with different severity levels in hydropower units. Firstly, four fine-grained indicators have been put forward to describe the severity of four kinds of common shafting faults. Secondly, a CNN fault diagnosis model is designed to extract high discriminative features and identify fault categories. Twelve kinds of shaft orbit samples were input into the model with the same size of 64×64.Through reaction of two convolution layers and two pooling layers in feature extraction module, the graphic features were extracted. The faults of hydropower units were identified after the fully connected layer and the SoftMax classifier. The experimental results show that the accuracy of the proposed approach for the fault with different severity is 98.33%, which is better than the traditional learning method. The proposed method retains the pattern information in the axis trajectory to the maximum extent and has the ability to accurately capture features of various severity faults.

Application of Data Enhancement Method Based on Generative Adversarial Networks for Soybean Leaf Disease Identification

Soybean leaf disease data collection is an expensive and time-consuming task. Convolutional neural network training requires a large amount of data, but traditional data enhancement methods (such as rotation, flipping, translation) are restricted by fixed rules and cannot generate images with diversity and variability. Aiming at the problem of the lack of soybean leaf disease data set, this study proposes a data enhancement method based on Generative Adversarial Networks (GANs) for soybean leaf disease identification. The method is based on a cyclic adversarial network and its discriminator uses dense connections. Strategies to reduce the size and computational complexity of the final model. Using a cyclic adversarial network to convert between healthy and diseased leaves, unsupervised learning can be performed, using limited images to learn disease characteristics, there by generating highly recognizable soybean leaf images. Synthesis images generated from GANs and original images are fed together as the model training set input and the recognition model for recognizing 9 types of soybean leaf images is obtained. An accuracy rate of 95.89% can be achieved on the verification set. Experimental results show that the generative adversarial network provided in this article can: Generate soybean leaf disease image data with high discriminative features, increase the size of the data set and provide a feasible solution for soybean leaf disease image data enhancement; as A regularization strategy to reduce over-fitting problems and improve the performance of the recognition model.

Fault Diagnosis Based on Attention Collaborative LSTM Networks for NPC Three-Level Inverters

To address the problem, it is difficult to extract fault features for neutral-point-clamped (NPC) three-level inverters under nonstationary conditions. This article proposes a multi-information feature fusion diagnosis method based on attention collaborative stacked long short-term memory (ASLSTM) neural networks. First, parallel structural stacked LSTM networks are constructed, which are used to automatically extract features from multisource time-series data. Then, the attention mechanism is applied to weight these features adaptively. Finally, the fault is identified by features that integrate multiple sources of information. In addition, the quantum particle swarm optimization (QPSO) algorithm is used to intelligently tune the hyperparameters of ASLSTM to improve the reasonableness of hyperparameter selection for the diagnostic model. The simulation results of the fault diagnosis of the NPC three-level inverter circuit show that the proposed method is able to extract high-discriminative features from raw data and has better diagnostic results under various conditions compared with other schemes.

Characterization of specific spatial functional connectivity difference in depression during sleep.

Depression is a common mental illness and a large number of researchers have been still devoted to exploring effective biomarkers for the identification of depression. Few researches have been conducted on functional connectivity (FC) during sleep in depression. In this paper, a novel depression characterization is proposed using specific spatial FC features of sleep electroencephalography (EEG). Overnight polysomnography recordings were obtained from 26 healthy individuals and 25 patients with depression. The weighted phase lag indexes (WPLIs) of four frequency bands and five sleep periods were obtained from 16 EEG channels. The high discriminative connections extracted via feature evaluation and the cross-within variation (CW)-the spatial feature constructed to characterize the different performances in inter- and intra-hemispheric FC based on WPLIs, were utilized to classify patients and normal controls. The results showed that enhanced average FC and spatial differences, higher inter-hemispheric FC and lower intra-hemispheric FC, were found in patients. Furthermore, abnormalities in the inter-hemispheric connections of the temporal lobe in the theta band should be important indicators of depression. Finally, both CW and high discriminative WPLI features performed well in depression screening and CW was more specific for characterizing abnormal cortical EEG performance of depression. Our work investigated and characterized the abnormalities in sleep cortical activity in patients with depression, and may provide potential biomarkers for assisting with depression identification and new insights into the understanding of pathological mechanisms in depression.

Fault diagnosis model for photovoltaic array using a dual-channels convolutional neural network with a feature selection structure

The effective fault diagnosis algorithm for the DC side photovoltaic (PV) array of a PV system (PVS) plays an important role in the operation efficiency and safety for PV power plants. But for fault diagnosis models it may fail to diagnose PV array (PVA) faults without detailed and quite fine fault features, especially line-line faults (LLF) occurring in the PVS that works under complex working conditions like low irradiance conditions and LLF with fault impedance. To address these challenges, this paper proposes a fault diagnosis scheme to diagnose different PVA faults using a proposed Dual-channel Convolutional Neural Network (DcCNN), which is able to automatically extract features and weight these features for fault classification. The important and fine features from the current and voltage electrical time series graph (ETSG) are extracted respectively by DcCNN in a double input way. Then, a proposed feature selection structure (FSS) is designed to improve the proposed fault diagnosis model capacity for diagnosing PVA faults under various conditions, including LLF, partial shading condition (PSC) and open circuit faults (OCF). Comparing to manually designed features, FSS not only helps DcCNN extract important features from PVA current and voltage automatically but also evaluates extracted features for further classification of DcCNN. Moreover, in the training stage, a proposed penalty is applied on DcCNN to constrain FSS, resulting in its sparse weight distribution. A comprehensive experiment based on a laboratory roof grid connected PVS is conducted. The results demonstrate the superior performance of the proposed approach compared with other algorithms as it can extract high-discriminative features from PVA current and voltage for different PVA faults, which is also effective on diagnosing LLF under low irradiance conditions and LLF with fault impedance.

CANet: Context Aware Network for Brain Glioma Segmentation

Automated segmentation of brain glioma plays an active role in diagnosis decision, progression monitoring and surgery planning. Based on deep neural networks, previous studies have shown promising technologies for brain glioma segmentation. However, these approaches lack powerful strategies to incorporate contextual information of tumor cells and their surrounding, which has been proven as a fundamental cue to deal with local ambiguity. In this work, we propose a novel approach named Context-Aware Network (CANet) for brain glioma segmentation. CANet captures high dimensional and discriminative features with contexts from both the convolutional space and feature interaction graphs. We further propose context guided attentive conditional random fields which can selectively aggregate features. We evaluate our method using publicly accessible brain glioma segmentation datasets BRATS2017, BRATS2018 and BRATS2019. The experimental results show that the proposed algorithm has better or competitive performance against several State-of-The-Art approaches under different segmentation metrics on the training and validation sets.

Mining Sports Articles using Cuckoo Search and Tabu Search with SMOTE Preprocessing Technique

Sentiment analysis is one of the most popular domains for natural language text classification, crucial for improving information extraction. However, massive data availability is one of the biggest problems for opinion mining due to accuracy considerations. Selecting high discriminative features from an opinion mining database is still an ongoing research topic. This study presents a two-stage heuristic feature selection method to classify sports articles using Tabu search and Cuckoo search via Levy flight. Levy flight is used to prevent the solution from being trapped at local optima. Comparative results on a benchmark dataset prove that our method shows significant improvements in the overall accuracy from 82.6% up to 89.5%.

Spectral–Spatial Joint Feature Extraction for Hyperspectral Image Based on High-Reliable Neighborhood Structure

In recent years, semisupervised spectral–spatial feature extraction (FE) methods for hyperspectral image (HSI) classification have shown promising performance by combining spectral–spatial information and label information. A problem that has not been addressed satisfactorily is that how to effectively and collaboratively use this abundant information contained in the hyperspectral data to exhibit better HSI classification performance. In this article, a novel FE method based on joint spectral–spatial information is proposed for HSI classification, which consists of the following steps. First, an effective re-expression for the original data is constructed by incorporating texture features extracted by extended multiattribute profiles with the original HSI. Thus, every pixel can be described by diverse and complementary information in the spectral–spatial domain. Then, the improved neighborhood preserving embedding (NPE) is proposed to establish a relatively accurate reconstruction model and mine high-reliable neighborhood structure from a global perspective by a new distance metric, which incorporates spectral bands, texture features, and geographical information simultaneously. Finally, the low-dimensional and high-discriminative features for HSI classification are obtained by combining the scatter matrices of local fisher discriminant analysis based on labeled samples and the improved NPE based on the whole data. Experimental results on three real-world HSI datasets show that the proposed method can effectively utilize both the label information and the spectral–spatial information, and hence, achieve much better classification performance compared to the conventional FE methods and some state-of-the-art spectral–spatial classification methods.

Multi-branch convolutional neural network with generalized shaft orbit for fault diagnosis of active magnetic bearing-rotor system

Fault diagnosis based on vibration signals in active magnetic bearing-rotor systems is an important research topic. However, it is difficult to obtain discriminative features to represent faults due to the nonlinear and non-stationary characteristics of the vibration signals and diverse sources of failures. Hence, this paper proposes a novel end-to-end learning mechanism of multi-sensor data fusion to learn fault representation based on the structural characteristics of active magnetic bearings. Taking the five displacement sensors of active magnetic bearing as signal sources, generalized shaft orbits are constructed and converted into discrete 2D images. Based these 2D images, a multi-branch convolutional neural network is designed to achieve high discriminative features and fault types. The experiments are performed on the rig supported by active magnetic bearings, and the effectiveness of the proposed algorithm is verified, proving it suitability in cases with changing rotating speeds and sample lengths.

Orthogonality Loss: Learning Discriminative Representations for Face Recognition

Convolutional neural networks have achieved excellent performance on face recognition (FR) by learning the high discriminative features with advanced loss functions. These improved loss functions share the similar idea for maximizing inter-class variance or minimizing intra-class variance. In this article, from a different perspective, we consider enlarging the inter-class variance by directly penalizing weight vectors of last fully connected layer, which represent the center of classes. To the end, we propose Orthogonality loss as an elegant penalty item appends to common classification loss to learn the discriminative representations. The main idea is that in order for weight vectors to be discriminative, it should be as close as possible to be orthogonal to each other in the vector space. More specifically, the optimization objective of Orthogonality loss is the first moment and second moment of cosine similarity of weight vectors. We performed the empirical studies through simulating the long-tail datasets to show the generalization ability of the proposed approach on long-tail distribution datasets. Further, extensive experiments on large-scale face recognition benchmarks including the Labeled Face in the Wild (LFW), the IARPA Janus Benchmark A (IJB-A), IJB-B, IJB-C, MegaFace Challenge 1 (MF1) and MS-Celeb-1M Low-shot Learning demonstrated that Orthogonality loss outperforms strong baselines, which showcases the extensive suitability and effectiveness of Orthogonality loss.