A new rotating machinery fault diagnosis method based on the Time Series Transformer

Fault diagnosis of rotating machinery is an important engineering problem. In recent years, fault diagnosis methods based on the Convolutional Neural Network (CNN) and Recurrent Neural Network (RNN) have been mature, but Transformer has not been widely used in the field of fault diagnosis. To address these deficiencies, a new method based on the Time Series Transformer (TST) is proposed to recognize the fault mode of bearings. In this paper, our contributions include: Firstly, we designed a tokens sequences generation method which can handle data in 1D format, namely time series tokenizer. Then, the TST combining time series tokenizer and Transformer was introduced. Furthermore, the test results on the given dataset show that the proposed method has better fault identification capability than the traditional CNN and RNN models. Secondly, through the experiments, the effect of structural hyperparameters such as subsequence length and embedding dimension on fault diagnosis performance, computational complexity and parameters number of the TST is analyzed in detail. The influence laws of some hyperparameters are obtained. Finally, via t-Distributed Stochastic Neighbor Embedding (t-SNE) dimensionality reduction method, the feature vectors in the embedding space are visualized. On this basis, the working pattern of TST has been explained to a certain extent. Moreover, by analyzing the distribution form of the feature vectors, we find that compared with the traditional CNN and RNN models, the feature vectors extracted by the method in this paper show the best intra-class compactness and inter-class separability. These results further demonstrate the effectiveness of the proposed method.

Real-space nanoimaging of THz polaritons in the topological insulator Bi2Se3

Dirac plasmon polaritons in topological insulators (TIs),light coupled to massless Dirac electrons, have been attracting a large amount of attention, both from a fundamental perspective and for potential terahertz (THz) photonic applications. Although THz polaritons have been observed by far-field THz spectroscopy on TI microstructures, real-space imaging of propagating THz polaritons in unstructured TIs has been elusive so far. Here, we show the very first spectroscopic THz near-field images of thin Bi2Se3 layers (prototypical TIs) revealing polaritons with up to 12 times increased momenta as compared to photons of the same energy and decay times of about 0.24 ps, yet short propagation lengths. From the near-field images we determine the polariton dispersions in layers from 120 to 25 nm thickness and perform a systematic theoretical dispersion analysis, showing that the observed polaritons can be explained only by the simultaneous coupling of THz radiation to Dirac carriers at the TI surfaces, massive bulk carriers and optical phonons. Our work does not only provide critical insights into the nature of THz polaritons in TIs, but also establishes instrumentation of unprecedented sensitivity for imaging of THz polaritons.

Independent Ethical Assessment of Text Classification Models: A Hate Speech Detection Case Study

An independent ethical assessment of an artificial intelligence system is an impartial examination of the systems development, deployment, and use in alignment with ethical values. System-level qualitative frameworks that describe high-level requirements and component-level quantitative metrics that measure individual ethical dimensions have been developed over the past few years. However, there exists a gap between the two, which hinders the execution of independent ethical assessments in practice. This study bridges this gap and designs a holistic independent ethical assessment process for a text classification model with a special focus on the task of hate speech detection. The assessment is further augmented with protected attributes mining and counterfactual-based analysis to enhance bias assessment. It covers assessments of technical performance, data bias, embedding bias, classification bias, and interpretability. The proposed process is demonstrated through an assessment of a deep hate speech detection model.

Analyzing Adversarial Robustness of Deep Neural Networks in Pixel Space: a Semantic Perspective

The vulnerability of deep neural networks to adversarial examples, which are crafted maliciously by modifying the inputs with imperceptible perturbations to misled the network produce incorrect outputs, reveals the lack of robustness and poses security concerns. Previous works study the adversarial robustness of image classifiers on image level and use all the pixel information in an image indiscriminately, lacking of exploration of regions with different semantic meanings in the pixel space of an image. In this work, we fill this gap and explore the pixel space of the adversarial image by proposing an algorithm to looking for possible perturbations pixel by pixel in different regions of the segmented image. The extensive experimental results on CIFAR-10 and ImageNet verify that searching for the modified pixel in only some pixels of an image can successfully launch the one-pixel adversarial attacks without requiring all the pixels of the entire image, and there exist multiple vulnerable points scattered in different regions of an image. We also demonstrate that the adversarial robustness of different regions on the image varies with the amount of semantic information contained.

Estimation of 3D Human Pose Using Prior Knowledge

Estimating three-dimensional human poses from the positions of two-dimensional joints has shown promising results.However, using two-dimensional joint coordinates as input loses more information than image-based approaches and results in ambiguity.In order to overcome this problem, we combine bone length and camera parameters with two-dimensional joint coordinates for input.This combination is more discriminative than the two-dimensional joint coordinates in that it can improve the accuracy of the models prediction depth and alleviate the ambiguity that comes from projecting three-dimensional coordinates into two-dimensional space. Furthermore, we introduce direction constraints which can better measure the difference between the ground truth and the output of the proposed model. The experimental results on the H36M show that the method performed better than other state-of-the-art three-dimensional human pose estimation approaches.

MOGAN: Morphologic-structure-aware Generative Learning from a Single Image

In most interactive image generation tasks, given regions of interest (ROI) by users, the generated results are expected to have adequate diversities in appearance while maintaining correct and reasonable structures in original images. Such tasks become more challenging if only limited data is available. Recently proposed generative models complete training based on only one image. They pay much attention to the monolithic feature of the sample while ignoring the actual semantic information of different objects inside the sample. As a result, for ROI-based generation tasks, they may produce inappropriate samples with excessive randomicity and without maintaining the related objects correct structures. To address this issue, this work introduces a MOrphologic-structure-aware Generative Adversarial Network named MOGAN that produces random samples with diverse appearances and reliable structures based on only one image. For training for ROI, we propose to utilize the data coming from the original image being augmented and bring in a novel module to transform such augmented data into knowledge containing both structures and appearances, thus enhancing the models comprehension of the sample. To learn the rest areas other than ROI, we employ binary masks to ensure the generation isolated from ROI. Finally, we set parallel and hierarchical branches of the mentioned learning process. Compared with other single image GAN schemes, our approach focuses on internal features including the maintenance of rational structures and variation on appearance. Experiments confirm a better capacity of our model on ROI-based image generation tasks than its competitive peers.

Machine Learning Percolation Model

Recent advances in machine learning have become increasingly popular in the applications of phase transitions and critical phenomena. By machine learning approaches, we try to identify the physical characteristics in the two-dimensional percolation model. To achieve this, we adopt Monte Carlo simulation to generate dataset at first, and then we employ several approaches to analyze the dataset. Four kinds of convolutional neural networks (CNNs), one variational autoencoder (VAE), one convolutional VAE (cVAE), one principal component analysis (PCA), and one $k$-means are used for identifying order parameter, the permeability, and the critical transition point. The former three kinds of CNNs can simulate the two order parameters and the permeability with high accuracy, and good extrapolating performance. The former two kinds of CNNs have high anti-noise ability. To validate the robustness of the former three kinds of CNNs, we also use the VAE and the cVAE to generate new percolating configurations to add perturbations into the raw configurations. We find that there is no difference by using the raw or the perturbed configurations to identify the physical characteristics, under the prerequisite of corresponding labels. In the case of lacking labels, we use unsupervised learning to detect the physical characteristics. The PCA, a classical unsupervised learning, performs well when identifying the permeability but fails to deduce order parameter. Hence, we apply the fourth kinds of CNNs with different preset thresholds, and identify a new order parameter and the critical transition point. Our findings indicate that the effectiveness of machine learning still needs to be evaluated in the applications of phase transitions and critical phenomena.

Dynamical evolution in a one-dimensional incommensurate lattice with $mathcal{PT}$ symmetry

We investigate the dynamical evolution of a parity-time ($mathcal{PT}$) symmetric extension of the Aubry-Andr{e} (AA) model, which exhibits the coincidence of a localization-delocalization transition point with a $mathcal{PT}$ symmetry breaking point. One can apply the evolution of the profile of the wave packet and the long-time survival probability to distinguish the localization regimes in the $mathcal{PT}$ symmetric AA model. The results of the mean displacement show that when the system is in the $mathcal{PT}$ symmetry unbroken regime, the wave-packet spreading is ballistic, which is different from that in the $mathcal{PT}$ symmetry broken regime. Furthermore, we discuss the distinctive features of the Loschmidt echo with the post-quench parameter being localized in different $mathcal{PT}$ symmetric regimes.

Information Restrain in Open Quantum Systems

Non-Hermitian skin effect of Liouvillian superoperators in quantum open systems can induce phenomena of non-trivial damping, known as chiral/helical damping. While non-Hermitian skin effect and chiral/helical damping occur only under open boundary condition, we propose an effect called information restrain which does not rely on boundary conditions. We demonstrate that information restrain is stable against disorder and is an intrinsic property of a type of open quantum systems or non-Hermitian system. Then we define the strength of information restrain $I_R$, which describes the ratio of different decay rates of signals strengthes along opposite propagation directions. Based on information restrain, We can provide a simple and elegant explanation of chiral and helical damping, and get the local maximum of relative particle number for periodical boundary system, consistent with numerical calculations. In terms of information restrain, we also illustrate the existence of correspondence between edge modes and damping modes and deduce that there are many chiral/helical transport properties in this information restrain class.

Improving Machine Reading Comprehension with Contextualized Commonsense Knowledge

In this paper, we aim to extract commonsense knowledge to improve machine reading comprehension. We propose to represent relations implicitly by situating structured knowledge in a context instead of relying on a pre-defined set of relations, and we call it contextualized knowledge. Each piece of contextualized knowledge consists of a pair of interrelated verbal and nonverbal messages extracted from a script and the scene in which they occur as context to implicitly represent the relation between the verbal and nonverbal messages, which are originally conveyed by different modalities within the script. We propose a two-stage fine-tuning strategy to use the large-scale weakly-labeled data based on a single type of contextualized knowledge and employ a teacher-student paradigm to inject multiple types of contextualized knowledge into a student machine reader. Experimental results demonstrate that our method outperforms a state-of-the-art baseline by a 4.3% improvement in accuracy on the machine reading comprehension dataset C^3, wherein most of the questions require unstated prior knowledge.