Abstract: In this paper, we propose a remote-sensing scene-classification method based on vision transformers. These types of networks, which are now recognized as state-of-the-art models in natural language processing, do not rely on convolution layers as in standard convolutional neural networks (CNNs). Instead, they use multihead attention mechanisms as the main building block to derive long-range contextual relation between pixels in images. In a first step, the images under analysis are divided into patches, then converted to sequence by flattening and embedding. To keep information about the position, embedding position is added to these patches. Then, the resulting sequence is fed to several multihead attention layers for generating the final representation. At the classification stage, the first token sequence is fed to a softmax classification layer. To boost the classification performance, we explore several data augmentation strategies to generate additional data for training. Moreover, we show experimentally that we can compress the network by pruning half of the layers while keeping competing classification accuracies. Experimental results conducted on different remote-sensing image datasets demonstrate the promising capability of the model compared to state-of-the-art methods. Specifically, Vision Transformer obtains an average classification accuracy of 98.49%, 95.86%, 95.56% and 93.83% on Merced, AID, Optimal31 and NWPU datasets, respectively. While the compressed version obtained by removing half of the multihead attention layers yields 97.90%, 94.27%, 95.30% and 93.05%, respectively.

Abstract: Learning discriminative features for Facial Expression Recognition (FER) in the wild using Convolutional Neural Networks (CNNs) is a non-trivial task due to the significant intra-class variations and inter-class similarities. Deep Metric Learning (DML) approaches such as center loss and its variants jointly optimized with softmax loss have been adopted in many FER methods to enhance the discriminative power of learned features in the embedding space. However, equally supervising all features with the metric learning method might include irrelevant features and ultimately degrade the generalization ability of the learning algorithm. We propose a Deep Attentive Center Loss (DACL) method to adaptively select a subset of significant feature elements for enhanced discrimination. The proposed DACL integrates an attention mechanism to estimate attention weights correlated with feature importance using the intermediate spatial feature maps in CNN as context. The estimated weights accommodate the sparse formulation of center loss to selectively achieve intra-class compactness and inter-class separation for the relevant information in the embedding space. An extensive study on two widely used wild FER datasets demonstrates the superiority of the proposed DACL method compared to state-of-the-art methods.

Abstract: Face forgery detection is raising ever-increasing interest in computer vision since facial manipulation technologies cause serious worries. Though recent works have reached sound achievements, there are still unignorable problems: a) learned features supervised by softmax loss are separable but not discriminative enough, since softmax loss does not explicitly encourage intra-class compactness and inter-class separability; and b) fixed filter banks and hand-crafted features are insufficient to capture forgery patterns of frequency from diverse inputs. To compensate for such limitations, a novel frequency-aware discriminative feature learning framework is proposed in this paper. Specifically, we design a novel single-center loss (SCL) that only compresses intra-class variations of natural faces while boosting inter-class differences in the embedding space. In such a case, the network can learn more discriminative features with less optimization difficulty. Besides, an adaptive frequency feature generation module is developed to mine frequency clues in a completely data-driven fashion. With the above two modules, the whole framework can learn more discriminative features in an end-to-end manner. Extensive experiments demonstrate the effectiveness and superiority of our framework on three versions of the FF++ dataset.

Abstract: Online class-incremental continual learning (CL) studies the problem of learning new classes continually from an online non-stationary data stream, intending to adapt to new data while mitigating catastrophic forgetting. While memory replay has shown promising results, the recency bias in online learning caused by the commonly used Softmax classifier remains an unsolved challenge. Although the Nearest-Class-Mean (NCM) classifier is significantly undervalued in the CL community, we demonstrate that it is a simple yet effective substitute for the Softmax classifier. It addresses the recency bias and avoids structural changes in the fully-connected layer for new classes. Moreover, we observe considerable and consistent performance gains when replacing the Softmax classifier with the NCM classifier for several state-of-the-art replay methods.To leverage the NCM classifier more effectively, data embeddings belonging to the same class should be clustered and well-separated from those with a different class label. To this end, we contribute Supervised Contrastive Replay (SCR), which explicitly encourages samples from the same class to cluster tightly in embedding space while pushing those of different classes further apart during replay-based training. Overall, we observe that our proposed SCR substantially reduces catastrophic forgetting and outperforms state-of-the-art CL methods by a significant margin on a variety of datasets.

Abstract: Learning-based optical flow estimation has been dominated with the pipeline of cost volume with convolutions for flow regression, which is inherently limited to local correlations and thus is hard to address the long-standing challenge of large displacements. To alleviate this, the state-of-the-art method, i.e., RAFT, gradually improves the quality of its predictions by producing a sequence of flow updates via a large number of iterative refinements, achieving remarkable performance but slowing down the inference speed. To enable both high accuracy and efficiency optical flow estimation, we completely revamp the dominating flow regression pipeline by reformulating optical flow as a global matching problem. Specifically, we propose a GMFlow framework, which consists of three main components: a customized Transformer for feature enhancement, a correlation and softmax layer for global feature matching, and a self-attention layer for flow propagation. Moreover, we further introduce a refinement step that reuses GMFlow at higher-resolutions for residual flow prediction. Our new framework outperforms 32-iteration RAFT's performance on the challenging Sintel benchmark, while using only one refinement and running faster, offering new possibilities for efficient and accurate optical flow estimation. Code will be available at https://github.com/haofeixu/gmflow.

Abstract: The current evaluation protocol of long-tailed visual recognition trains the classification model on the long-tailed source label distribution and evaluates its performance on the uniform target label distribution. Such protocol has questionable practicality since the target may also be long-tailed. Therefore, we formulate long-tailed visual recognition as a label shift problem where the tar-get and source label distributions are different. One of the significant hurdles in dealing with the label shift problem is the entanglement between the source label distribution and the model prediction. In this paper, we focus on disentangling the source label distribution from the model prediction. We first introduce a simple but over-looked baseline method that matches the target label distribution by post-processing the model prediction trained by the cross-entropy loss and the Softmax function. Al-though this method surpasses state-of-the-art methods on benchmark datasets, it can be further improved by directly disentangling the source label distribution from the model prediction in the training phase. Thus, we propose a novel method, LAbel distribution DisEntangling (LADE) loss based on the optimal bound of Donsker-Varadhan representation. LADE achieves state-of-the-art performance on benchmark datasets such as CIFAR-100-LT, Places-LT, ImageNet-LT, and iNaturalist 2018. Moreover, LADE out-performs existing methods on various shifted target label distributions, showing the general adaptability of our pro-posed method.

Abstract: Learning discriminative face features plays a major role in building high-performing face recognition models. The recent state-of-the-art face recognition solutions proposed to incorporate a fixed penalty margin on commonly used classification loss function, softmax loss, in the normalized hypersphere to increase the discriminative power of face recognition models, by minimizing the intra-class variation and maximizing the inter-class variation. Marginal softmax losses, such as ArcFace and CosFace, assume that the geodesic distance between and within the different identities can be equally learned using a fixed margin. However, such a learning objective is not realistic for real data with inconsistent inter-and intra-class variation, which might limit the discriminative and generalizability of the face recognition model. In this paper, we relax the fixed margin constrain by proposing elastic margin loss (ElasticFace) that allows flexibility in the push for class separability. The main idea is to utilize random margin values drawn from a normal distribution in each training iteration. This aims at giving the margin chances to extract and retract to allow space for flexible class separability learning. We demonstrate the superiority of our elastic margin loss over ArcFace and CosFace losses, using the same geometric transformation, on a large set of mainstream benchmarks. From a wider perspective, our ElasticFace has advanced the state-of-the-art face recognition performance on six out of nine mainstream benchmarks.

Abstract: Deep learning techniques have been widely applied to hyperspectral image (HSI) classification and have achieved great success. However, the deep neural network model has a large parameter space and requires a large number of labeled data. Deep learning methods for HSI classification usually follow a patchwise learning framework. Recently, a fast patch-free global learning (FPGA) architecture was proposed for HSI classification according to global spatial context information. However, FPGA has difficulty in extracting the most discriminative features when the sample data are imbalanced. In this article, a spectral-spatial-dependent global learning (SSDGL) framework based on the global convolutional long short-term memory (GCL) and global joint attention mechanism (GJAM) is proposed for insufficient and imbalanced HSI classification. In SSDGL, the hierarchically balanced (H-B) sampling strategy and the weighted softmax loss are proposed to address the imbalanced sample problem. To effectively distinguish similar spectral characteristics of land cover types, the GCL module is introduced to extract the long short-term dependency of spectral features. To learn the most discriminative feature representations, the GJAM module is proposed to extract attention areas. The experimental results obtained with three public HSI datasets show that the SSDGL has powerful performance in insufficient and imbalanced sample problems and is superior to other state-of-the-art methods.

Abstract: Most works on person re-identification (ReID) take advantage of large backbone networks such as ResNet, which are designed for image classification instead of ReID, for feature extraction. However, these backbones may not be computationally efficient or the most suitable architectures for ReID. In this work, we aim to design a lightweight and suitable network for ReID. We propose a novel search space called Combined Depth Space (CDS), based on which we search for an efficient network architecture, which we call CDNet, via a differentiable architecture search algorithm. Through the use of the combined basic building blocks in CDS, CDNet tends to focus on combined pattern information that is typically found in images of pedestrians. We then propose a low-cost search strategy named the Top-k Sample Search strategy to make full use of the search space and avoid trapping in local optimal result. Furthermore, an effective Fine-grained Balance Neck (FBLNeck), which is removable at the inference time, is presented to balance the effects of triplet loss and softmax loss during the training process. Extensive experiments show that our CDNet (~1.8 M parameters) has comparable performance with state-of-the-art lightweight networks.

Abstract: Fault diagnosis of rolling bearings is an essential process for improving the reliability and safety of the rotating machinery. It is always a major challenge to ensure fault diagnosis accuracy in particular under severe working conditions. In this article, a deep adversarial domain adaptation (DADA) model is proposed for rolling bearing fault diagnosis. This model constructs an adversarial adaptation network to solve the commonly encountered problem in numerous real applications: the source domain and the target domain are inconsistent in their distribution. First, a deep stack autoencoder (DSAE) is combined with representative feature learning for dimensionality reduction, and such a combination provides an unsupervised learning method to effectively acquire fault features. Meanwhile, domain adaptation and recognition classification are implemented using a Softmax classifier to augment classification accuracy. Second, the effects of the number of hidden layers in the stack autoencoder network, the number of neurons in each hidden layer, and the hyperparameters of the proposed fault diagnosis algorithm are analyzed. Third, comprehensive analysis is performed on real data to validate the performance of the proposed method; the experimental results demonstrate that the new method outperforms the existing machine learning and deep learning methods, in terms of classification accuracy and generalization ability.

Abstract: As a deep learning model, a deep belief network (DBN) consists of multiple restricted Boltzmann machines (RBMs). Based on DBN, many intelligent fault diagnosis methods are proposed. However, these methods seldom considered the appearance of new working conditions during the operation of real machines. Varying working conditions lead to a change of feature distributions and finally result in low diagnosis accuracies. Therefore, we propose a distribution-invariant DBN (DIDBN) to learn distribution-invariant features directly from raw vibration data. DIDBN consists of a locally connected RBM (LCRBM) layer, a fully connected RBM layer, and an RBM layer with a mean discrepancy maximum (MDM-RBM). The LCRBM layer is designed with a locally connected structure. By proposing MDM, the MDM-RBM layer is able to obtain features that have close distributions under varying working conditions. Followed by a softmax classifier, DIDBN is able to recognize faults. The proposed method is applied to two diagnosis cases. Results verify that DIDBN is able to learn distribution-invariant features and achieve higher diagnosis accuracies than recently proposed methods. Moreover, the reason why DIDBN is able to learn distribution-invariant features is explained by visualizing the feature learning process.

Abstract: Recently, recommender systems are widely used on various platforms in real world to provide personalized recommendations. However, sparsity is a tough problem in a Collaborate Filtering (CF) recommender system as it always leads to the over-fitting problem. This paper proposes a Model-based Collaborate Filtering Algorithm Based on Stacked AutoEncoder (MCFSAE) to overcome the sparsity problem. In the MCFSAE model, we first convert the rating matrix into a high-dimensional classification dataset with a size equal to the number of ratings. As the number of ratings is usually large scale, the classification performance can be guaranteed. Since the obtained classification dataset is high dimensional, we then utilize Stacked AutoEncoder, which is a good nonlinear feature reduction model, to obtain a high-level low-dimensional feature presentation. Finally, a softmax classification model is used to predict the unknown ratings based on the high-level features. Extensive experiments on EachMovie and MovieLens datasets are conducted to compare the proposed MCFSAE model with other SOTA CF models. Experimental results show that MCFSAE performs better than other CF models, especially when the rating matrix is sparse.

Abstract: Transformers are state-of-the-art models for a variety of sequence modeling tasks. At their core is an attention function which models pairwise interactions between the inputs at every timestep. While attention is powerful, it does not scale efficiently to long sequences due to its quadratic time and space complexity in the sequence length. We propose RFA, a linear time and space attention that uses random feature methods to approximate the softmax function, and explore its applications in transformers. RFA offers a straightforward way of learning with recency bias through an optional gating mechanism and can be used as a drop-in replacement for conventional softmax attention. Experiments on language modeling and machine translation demonstrate that RFA achieves similar or better performance compared to strong transformer baselines. In the machine translation experiment, RFA decodes twice as fast as a vanilla transformer. Compared to existing efficient transformer variants, RFA is competitive in terms of both accuracy and efficiency on three long text classification datasets. Our analysis shows that RFA’s efficiency gains are especially notable on long sequences, suggesting that RFA will be particularly useful in tasks that require working with large inputs, fast decoding speed, or low memory footprints.

Abstract: Graph representation learning aims to embed each vertex in a graph into a low-dimensional vector space. Existing graph representation learning methods can be classified into two categories: generative models that learn the underlying connectivity distribution in a graph, and discriminative models that predict the probability of edge between a pair of vertices. In this paper, we propose GraphGAN , an innovative graph representation learning framework unifying the above two classes of methods, in which the generative and the discriminative model play a game-theoretical minimax game. Specifically, for a given vertex, the generative model tries to fit its underlying true connectivity distribution over all other vertices and produces “fake” samples to fool the discriminative model, while the discriminative model tries to detect whether the sampled vertex is from ground truth or generated by the generative model. With the competition between these two models, both of them can alternately and iteratively boost their performance. Moreover, we propose a novel graph softmax as the implementation of the generative model to overcome the limitations of traditional softmax function, which can be proven satisfying desirable properties of normalization , graph structure awareness , and computational efficiency . Through extensive experiments on real-world datasets, we demonstrate that GraphGAN achieves substantial gains in a variety of applications, including graph reconstruction, link prediction, node classification, recommendation, and visualization, over state-of-the-art baselines.

Abstract: Federated Learning (FL) aims to generate a global shared model via collaborating decentralized clients with privacy considerations. Unlike standard distributed optimization, FL takes multiple optimization steps on local clients and then aggregates the model updates via a parameter server. Although this significantly reduces communication costs, the non-iid property across heterogeneous devices could make the local update diverge a lot, posing a fundamental challenge to aggregation. In this paper, we focus on a special kind of non-iid scene, i.e., label distribution skew, where each client can only access a partial set of the whole class set. Considering top layers of neural networks are more task-specific, we advocate that the last classification layer is more vulnerable to the shift of label distribution. Hence, we in-depth study the classifier layer and point out that the standard softmax will encounter several problems caused by missing classes. As an alternative, we propose "Restricted Softmax" to limit the update of missing classes' weights during the local procedure. Our proposed FedRS is very easy to implement with only a few lines of code. We investigate our methods on both public datasets and a real-world service awareness application. Abundant experimental results verify the superiorities of our methods.

Abstract: Intelligent fault diagnosis based on deep neural networks and big data has been an attractive field and shows great prospects for applications. However, applications in practice face following problems. (1) Unexpected and unseen faults of machinery in real environment may be encountered. (2) Large collections of healthy condition samples and few fault condition samples result in the imbalanced distribution of machinery health conditions. This paper proposes a novel deep metric learning model, where machinery condition is classified by retrieving similarities. The trained deep metric learning model can learn and recognize new faults quickly and easily to address the first problem. As core of deep metric learning, a novel loss function called normalized softmax loss with adaptive angle margin (NSL-AAM) is developed for second problem. NSL-AAM can supervise neural networks learning imbalanced data without altering the original data distribution. Experiments for balanced and imbalanced fault diagnosis are conducted to verify the ability of the proposed model for fault diagnosis. The results demonstrate that the proposed model can not only extract more distinctive features automatically, but also balance the representation of both the majority and minority classes. Furthermore, the results of experiments for diagnosing new faults are reported, which proves the capability of the trained model for open-set classification.

Abstract: The key to intelligent fault diagnosis is to find relevant characteristics with the capability of representing different types of faults. However, the engineering problem is that a few simple empirical features (EFs) cannot obtain high classification accuracy, and complex feature engineering requires strong professional knowledge, which leads to limited applications on a general scale. In addition, intelligent feature extraction and classification methods without prior knowledge cannot guarantee that the model learned the general features used for classification, and its robustness and generalization are not strong when the objects with low-quality training data. Therefore, a fusion method of combining EFs and adaptive features extracted by a deep neural network is proposed. In this method, simple EFs that only need a few professional knowledge are adopted to realize general feature extraction and, hence, maintain the robustness of the model. A modified neural network structure (LiftingNet) is proposed to achieve the adaptive extraction of hidden features for specific objects, for realizing the high precision of bearing fault classification. In order to realize the fusing of EFs and adaptive features, XGBoost is utilized as the final classifier instead of common softmax. The feasibility and validity of the proposed method are verified by two data sets collected from motor bearings at stable work conditions. The experimental results show that the classification accuracy generated by the proposed method is improved. It also can maintain the robust performance on data sets even with various noises.

Abstract: We introduce Performers, Transformer architectures which can estimate regular (softmax) full-rank-attention Transformers with provable accuracy, but using only linear (as opposed to quadratic) space and time complexity, without relying on any priors such as sparsity or low-rankness. To approximate softmax attention-kernels, Performers use a novel Fast Attention Via positive Orthogonal Random features approach (FAVOR+), which may be of independent interest for scalable kernel methods. FAVOR+ can also be used to efficiently model kernelizable attention mechanisms beyond softmax. This representational power is crucial to accurately compare softmax with other kernels for the first time on large-scale tasks, beyond the reach of regular Transformers, and investigate optimal attention-kernels. Performers are linear architectures fully compatible with regular Transformers and with strong theoretical guarantees: unbiased or nearly-unbiased estimation of the attention matrix, uniform convergence and low estimation variance. We tested Performers on a rich set of tasks stretching from pixel-prediction through text models to protein sequence modeling. We demonstrate competitive results with other examined efficient sparse and dense attention methods, showcasing effectiveness of the novel attention-learning paradigm leveraged by Performers.

Abstract: Vision-based human action recognition (HAR) is a hot topic of research from the decade due to a few popular applications such as visual surveillance and robotics. For correct action recognition, various local and global points are requires known as features. These features modified during the variation in human movement. But due to a bit change in several human actions, the features of these actions are mixed that degrade the recognition performance. In this article, we design a new 26-layered Convolutional Neural Network (CNN) architecture for accurate complex action recognition. The features are extracted from the global average pooling layer and fully connected (FC) layer, and fused by a proposed high entropy-based approach. Further, we propose a feature selection method name Poisson distribution along with Univariate Measures (PDaUM). Few of fused CNN features are irrelevant, and few of them are redundant that makes the incorrect prediction among complex human actions. Therefore, the proposed PDaUM based approach selects only the strongest features that later passed to the Extreme Learning Machine (ELM) and Softmax for final recognition. Four datasets are using for experimental analysis - HMDB51 (51 classes), UCF Sports (10 classes), KTH (6 classes), and Weizmann (10 classes). On these datasets, the ELM classifier gives an improved performance as compared to a Softmax classifier. The achieved accuracy on each dataset is 81.4%, 99.2%, 98.3%, and 98.7%, respectively. Comparison with existing techniques, it is shown that the proposed architecture gives better performance in terms of accuracy and testing time.

Abstract: The current COVID-19 pandemic threatens human life, health, and productivity. AI plays an essential role in COVID-19 case classification as we can apply machine learning models on COVID-19 case data to predict infectious cases and recovery rates using chest x-ray. Accessing patient's private data violates patient privacy and traditional machine learning model requires accessing or transferring whole data to train the model. In recent years, there has been increasing interest in federated machine learning, as it provides an effective solution for data privacy, centralized computation, and high computation power. In this paper, we studied the efficacy of federated learning versus traditional learning by developing two machine learning models (a federated learning model and a traditional machine learning model)using Keras and TensorFlow federated, we used a descriptive dataset and chest x-ray (CXR) images from COVID-19 patients. During the model training stage, we tried to identify which factors affect model prediction accuracy and loss like activation function, model optimizer, learning rate, number of rounds, and data Size, we kept recording and plotting the model loss and prediction accuracy per each training round, to identify which factors affect the model performance, and we found that softmax activation function and SGD optimizer give better prediction accuracy and loss, changing the number of rounds and learning rate has slightly effect on model prediction accuracy and prediction loss but increasing the data size did not have any effect on model prediction accuracy and prediction loss. finally, we build a comparison between the proposed models' loss, accuracy, and performance speed, the results demonstrate that the federated machine learning model has a better prediction accuracy and loss but higher performance time than the traditional machine learning model.

Abstract: Since electroencephalogram (EEG) signals can truly reflect human emotional state, emotion recognition based on EEG has turned into a critical branch in the field of artificial intelligence. Aiming at the disparity of EEG signals in various emotional states, we propose a new deep learning model named three-dimension convolution attention neural network (3DCANN) for EEG emotion recognition in this paper. The 3DCANN model is composed of spatio-temporal feature extraction module and EEG channel attention weight learning module, which can extract the dynamic relation well among multi-channel EEG signals and the internal spatial relation of multi-channel EEG signals during continuous time period. In this model, the spatio-temporal features are fused with the weights of dual attention learning, and the fused features are input into softmax classifier for emotion classification. In addition, we utilize SJTU Emotion EEG Dataset (SEED) to appraise the feasibility and effectiveness of the proposed algorithm. Finally, experimental results display that the 3DCANN method has superior performance over the state-of-the-art models in EEG emotion recognition.

Abstract: Evaluating the log-sum-exp function or the softmax function is a key step in many modern data science algorithms, notably in inference and classification. Because of the exponentials that these functions contain, the evaluation is prone to overflow and underflow, especially in low precision arithmetic. Software implementations commonly use alternative formulas that avoid overflow and reduce the chance of harmful underflow, employing a shift or another rewriting. Although mathematically equivalent, these variants behave differently in floating-point arithmetic ew{and shifting can introduce subtractive cancellation}. We give rounding error analyses of different evaluation algorithms and interpret the error bounds using condition numbers for the functions. We conclude, based on the analysis and numerical experiments, that the shifted formulas are of similar accuracy to the unshifted ones, so can safely be used, but that a division-free variant of softmax can suffer from loss of accuracy.

Abstract: Deep learning (DL) has exhibited huge potentials for hyperspectral image (HSI) classification due to its powerful nonlinear modeling and end-to-end optimization characteristics. Although the superior performance of DL-based methods has been witnessed, some limitations can still be found. On the one hand, existing DL frameworks usually resorted to first-order statistical features, whereas they rarely considered second-order or higher order statistical features. On the other hand, the optimization of complex hyperparameters (e.g., the layer number and convolutional kernel size) is time-consuming and a very tough task, making the designed DL framework unexplainable. To overcome these challenges, we propose a novel attention-based second-order pooling network (A-SPN). First, a first-order feature operator is designed to model the spectral–spatial information of HSI. Second, an attention-based second-order pooling (A-SOP) operator is designed to model discriminative and representative features. Finally, a fully connected layer with softmax loss is used for classification. The proposed framework can obtain second-order statistical features in an end-to-end manner. In addition, A-SPN is free of complex hyperparameters tuning, making it more explainable and easily equipped for classification tasks. Experimental results based on three common hyperspectral data sets demonstrate that A-SPN outperforms other traditional and state-of-the-art DL-based HSI classification methods in terms of generalization performance with limited training samples, classification accuracy, convergence rate, and computational complexity.

Abstract: Deep face recognition has achieved remarkable improvements due to the introduction of margin-based softmax loss, in which the prototype stored in the last linear layer represents the center of each class. In these methods, training samples are enforced to be close to positive prototypes and far apart from negative prototypes by a clear margin. However, we argue that prototype learning only employs sample-to-prototype comparisons without considering sample-to-sample comparisons during training and the low loss value gives us an illusion of perfect feature embedding, impeding the further exploration of SGD. To this end, we propose Variational Prototype Learning (VPL), which represents every class as a distribution instead of a point in the latent space. By identifying the slow feature drift phenomenon, we directly inject memorized features into prototypes to approximate variational prototype sampling. The proposed VPL can simulate sample-to-sample comparisons within the classification framework, encouraging the SGD solver to be more exploratory, while boosting performance. Moreover, VPL is conceptually simple, easy to implement, computationally efficient and memory saving. We present extensive experimental results on popular benchmarks, which demonstrate the superiority of the proposed VPL method over the state-of-the-art competitors.

Abstract: Objective: Filter bank canonical correlation analysis (FBCCA) is a widely-used classification approach implemented in steady-state visual evoked potential (SSVEP)–based brain-computer interfaces (BCIs). However, conventional detection algorithms for SSVEP recognition problems, including the FBCCA, were usually based on 'fixed window' strategy. That's to say, these algorithms always analyze data with fixed length. This study devoted to enhance the performance of SSVEP-based BCIs by designing a new dynamic window strategy which automatically finds an optimal data length to achieve higher information transfer rate (ITR). Approach: The main purpose of 'dynamic window' is to minimize the required data length while maintaining high accuracy. This study projected the correlation coefficients of FBCCA into probability space by softmax function and built a hypothesis testing model, which took risk function as evaluation of classification result's 'credibility'. In order to evaluate the superiority of this approach, FBCCA with fixed data length (FBCCA-FW) and spatial temporal equalization dynamic window (STE-DW) were implemented for comparison. Main results: Fourteen healthy subjects' results were concluded by a 40-target online SSVEP-based BCI speller system. The results suggest that this proposed approach significantly outperforms STE-DW and FBCCA-FW in terms of accuracy and ITR. Significance: By incorporating the fundamental ideas of FBCCA and dynamic window strategy, this study proposed a new training-free dynamical optimization algorithm, which significantly improved the performance of online SSVEP-based BCI systems.

Abstract: Fault diagnosis plays an indispensable role in prognostics and health management of rotating machines. In recent years, intelligent fault diagnosis methods based on domain adaptation technology have attracted the attention of researchers. However, a more extensive application scenario of fault diagnosis − partial domain adaptation (PDA) − has not been well-resolved. In this article, for the first time, a novel stacked auto-encoder based partial adversarial domain adaptation (SPADA) model is proposed to solve the fault diagnosis problem in PDA situations. Two deep stack auto-encoders are first designed to extract representative features from the training data (source domain) and test data (target domain), respectively. Then, a weighted classifier based on Softmax is used to weight the features from the source and target domains. Meanwhile, another domain discriminator and label predictor using the Softmax classifier are adopted to simultaneously implement domain adaptation and fault diagnosis. Comprehensive analysis is performed on real data to test the performance of the SPADA model and detailed comparisons are provided; the extensive experimental results show that the diagnosis performance of SPADA outperforms the existing deep learning and domain adaptation methods in dealing with the PDA problem.