Noise Vector Research Articles

Robust Bearing-Only Localization Using Total Least Absolute Residuals Optimization

Robust techniques critically improve bearing-only target localization when the relevant measurements are being corrupted by impulsive noise. Resistance to isolated gross errors refers to the conventional least absolute residual (LAR) method, and its estimate can be determined by linear programming when pseudolinear equations are set. The LAR approach, however, cannot reduce the bias attributed to the correlation between system matrices and noise vectors. In the present study, perturbations are introduced into the elements of the system matrix and the data vector simultaneously, and the total optimization problem is formulated based on least absolute deviations. Subsequently, an equivalent form of total least absolute residuals (TLAR) is obtained, and an algorithm is developed to calculate the robust estimate by dual ascent algorithms. Moreover, the performance of the proposed method is verified through the numerical simulations by using two types of localization geometries, i.e., random and linear. As revealed from the results, the TLAR algorithm is capable of exhibiting significantly higher localization accuracy as compared with the LAR method.

Forecasting the Heat Load of Residential Buildings with Heat Metering Based on CEEMDAN-SVR

The energy demand of the district heating system (DHS) occupies an important part in urban energy consumption, which has a great impact on the energy security and environmental protection of a city. With the gradual improvement of people’s economic conditions, different groups of people now have different demands for thermal energy for their comfort. Hence, heat metering has emerged as an imperative for billing purposes and sustainable management of energy consumption. Therefore, forecasting the heat load of buildings with heat metering on the demand side is an important management strategy for DHSs to meet end-users’ needs and maintain energy-saving regulations and safe operation. However, the non-linear and non-stationary characteristics of buildings’ heat load make it difficult to predict consumption patterns accurately, thereby limiting the capacity of the DHS to deliver on its statutory functions satisfactorily. A novel ensemble prediction model is proposed to resolve this problem, which integrates the advantages of Complete Ensemble Empirical Mode Decomposition with Adaptive Noise (CEEMDAN) and support vector regression (SVR), called CEEMDAN-SVR in this paper. The proposed CEEMDAN-SVR algorithm is designed to automatically decompose the intrinsic mode according to the characteristics of heat load data to ensure an accurate representation of heat load patterns on multiple time scales. It will also be useful for developing an accurate prediction model for the buildings’ heat load. In formulating the CEEMDAN-SVR model, the heat load data of three different buildings in Xingtai City were acquired during the heating season of 2019–2020 and employed to conduct detailed comparative analysis with modern algorithms, such as extreme tree regression (ETR), forest tree regression (FTR), gradient boosting regression (GBR), support vector regression (SVR, with linear, poly, radial basis function (RBF) kernel), multi-layer perception (MLP) and EMD-SVR. Experimental results reveal that the performance of the proposed CEEMDAN-SVR model is better than the existing modern algorithms and it is, therefore, more suitable for modeling heat load forecasting.

INS/UWB-Based Quadrotor Localization Under Colored Measurement Noise

Data fusion in the quadrotor integrated navigation system combining information from the inertial navigation system (INS) and ultra wide band (UWB)-based system is provided in the presence of the UWB-induced colored measurement noise (CMN). To improve the localization performance under the CMN, we modify the fusion Kalman filter (KF) to be cKF and the unbiased finite impulse response (UFIR) filter to be cUFIR filter. The cKF and cUFIR filter are derived using measurement differencing and by de-correlating noise vectors. The original and modified filters are tested in the best and worst cases of tuning parameters. Based on test measurements without the ground truth, it is shown that the cKF improves the performance in the best case, but loses to the KF in the worst case due to a higher sensitivity to errors in the noise covariances. It is also shown that the cUFIR filter is more robust and thus more suitable than the cKF and KF for quadrotor localization under the industrial conditions.

Text to image synthesis using multi-generator text conditioned generative adversarial networks

Recently, Generative Adversarial Network(GAN) has been the most mainstream technology in the task of Text to Image. However, the vanilla deep neural networks tend to approximate continuous mappings in real generation tasks rather than discontinuous mappings with discrete points. When training on datasets with multiple types, GAN fails to synthesize diverse images, which we call as mode collapse. To deal with it, we propose the Multi-generator Text Conditioned Generative Adversarial Network (MTC-GAN) in this paper. Textual description of real images is embedded on the noise vector as a constraint. Based on Deep Convolutional Generative Adversarial Networks(DCGAN), multiple generators are incorporated to capture high probability among the target distribution. To identify the generated fake sample from a particular generator, the discriminator must enforce multiple generators to have different identifiable modes. The method based on global constraints can make the generated images more diverse. Multiple generators can improve the particular functional shape of the discriminators indirectly, which should make the GAN more stable when trained in high dimensional spaces. The experimental results on the standard dataset demonstrate the good performance of the proposed method. The problem of mode collapse can be improved, and the generated samples can be more diverse.

GANalyze: Toward visual definitions of cognitive image properties

We introduce a framework that uses Generative Adversarial Networks (GANs) to study cognitive image properties (e.g., memorability, aesthetics, valence). Often, we do not have concrete visual definitions of what these properties entail. Starting from input noise, GANs generate a manifold of natural-looking images with fine-grained differences in their visual attributes. By navigating this manifold, we can visualize what it looks like for a particular GAN-image to become more (less) memorable. Specifically, we trained a Transformer module to learn along which direction to move a BigGAN-image’s corresponding noise vector in order to increase (or decrease) its memorability. Memorability was assessed by an off-the-shelf Assessor (MemNet). After training, we generated a test set of 1.5K “seed images”, each with four “clone images”: two modified to be more memorable (one and two “steps” forward along the learned direction) and two to be less memorable (one and two steps backward; examples in Supplemental). The assessed memorability significantly increased when stepping along the learned direction (β = 0.68, p < 0.001), suggesting training was successful. Through a behavioral repeat-detection memory experiment, we verified that our method’s manipulations indeed causally affect human memory performance (β = 1.92, p < 0.001). The seeds and their clones (i.e., "visual definitions") surfaced candidate image properties (e.g., “object size”, “colorfulness”) that may underlie memorability and were previously overlooked. These candidates correlated with the learned memorability direction. We furthermore demonstrate that stepping along a learned “object size” direction indeed increases human memorability, though less strongly (β = 0.11, p < 0.001). This showcases how the individual, causal effects of a candidate can be studied further using the same framework. Finally, we find that by substituting the Assessor, our framework can also provide visual definitions for aesthetics (β = 0.72, p < 0.001) and emotional valence (β = 0.44, p < 0.001).

Denoising Sentinel-1 Extra-Wide Mode Cross-Polarization Images Over Sea Ice

Sentinel-1 (S1) extra-wide (EW) swath data in cross-polarization (horizontal–vertical, HV or vertical–horizontal, VH) are strongly affected by the scalloping effect and thermal noise, particularly over areas with weak backscattered signals, such as sea surfaces. Although noise vectors in both the azimuth and range directions are provided in the standard S1 EW data for subtraction, the residual thermal noise still significantly affects sea ice detection by the EW data. In this article, we improve the denoising method developed in previous studies to remove the additive noise for the S1 EW data in cross-polarization. Furthermore, we propose a new method for eliminating the residual noise (i.e., multiplicative noise) at the subswath boundaries of the EW data, which cannot be well processed by simply subtracting the reconstructed 2-D noise field. The proposed method of removing both the additive and multiplicative noise was applied to EW HV-polarized images processed using different Instrument Processing Facility (IPF) versions. The results suggest that the proposed algorithm significantly improves the quality of EW HV-polarized images under various sea ice conditions and sea states in the marginal ice zone (MIZ) of the Arctic. This is of great support for the utilization of cross-polarization synthetic aperture radar (SAR) images in wide swaths for intensive sea ice monitoring in polar regions.

Generating Video from Images using GANs

Generative adversarial networks are a category of neural networks used extensively for the generation of a wide range of content. The generative models are trained through an adversarial process that offers a lot of potential in the world of deep learning. GANs are a popular approach to generate new data from random noise vector that are similar or have the same distribution as that in the training data set. The Generative Adversarial Networks (GANs) approach has been proposed to generate more realistic images. An extension of GANs is the conditional GANs which allows the model to condition external information. Conditional GANs have seen increasing uses and more implications than ever. We also propose a new framework for estimating generative models via an adversarial process, in which we simultaneously train two models, a generative model G that captures the data distribution, and a discriminative model D that estimates the probability that a sample came from the training data rather than G. Our work aims at highlighting the uses of conditional GANs specifically with Generating images. We present some of the use cases of conditional GANs with images specifically in video generation.

Shale Digital Core Image Generation Based on Generative Adversarial Networks

Abstract Stochastic reconstruction of digital core images is a vital part of digital core physics analysis, aiming to generate representative microstructure samples for sampling and uncertainty quantification analysis. This paper proposes a novel reconstruction method of the digital core of shale based on generative adversarial networks (GANs) with powerful capabilities of the generation of samples. GANs are a series of unsupervised generative artificial intelligence models that take the noise vector as an input. In this paper, the GANs with a generative and a discriminative network are created respectively, and the shale image with 45 nm/pixel preprocessed by the three-value-segmentation method is used as training samples. The generative network is used to learn the distribution of real training samples, and the discriminative network is used to distinguish real samples from synthetic ones. Finally, realistic digital core samples of shale are successfully reconstructed through the adversarial training process. We used the Fréchet inception distance (FID) and Kernel inception distance (KID) to evaluate the ability of GANs to generate real digital core samples of shale. The comparison of the morphological characteristics between them, such as the ratio of organic matter and specific surface area of organic matter, indicates that real and reconstructed samples are highly close. The results show that deep convolutional generative adversarial networks with full convolution properties can reconstruct digital core samples of shale effectively. Therefore, compared with the classical methods of reconstruction, the new reconstruction method is more promising.

Man-In-The-Middle Attack against Certain Authentication Protocols Revisited: Insights into the Approach and Performances Re-Evaluation

We address a class of authentication protocols called “HB” ones and the man-in-the-middle (MIM) attack, reported at the ASIACRYPT conference, called OOV-MIM (Ouafi-Overbeck-Vaudenay MIM). Analysis of the considered attack and its systematic experimental evaluation are given. It is shown that the main component of OOV-MIM, the algorithm for measuring the Hamming weight of noise vectors, outputs incorrect results as a consequence of the employed approximation of the probability distributions. The analysis reveals that, practically, the only scenario in which the OOV-MIM attack is effective is the one in which two incorrect estimations produced by the algorithm for measuring the Hamming weight, when coupled, give the correct result. This paper provides additional insights into the OOV-MIM and corrected claims about the performance/complexity showing that the performances of the considered attack have been overestimated, i.e., that the complexity of the attack has been underestimated. Particularly, the analysis points out the reasons for the incorrect claims and to the components of the attack that do not work as expected.

HybridGAN: hybrid generative adversarial networks for MR image synthesis

In this paper, we propose HybridGAN – a new medical MR image synthesis methods via generative adversarial learning. Specifically, our synthesizer generates MRI data in a sequential manner: first in order to improve the robustness of image synthesis, an input full-size real MR image is divided into an array of sub-images. Then, to avoid overfitting limited MRI encodings, these sub-images and an unlimited amount of random latent noise vectors become the input of automatic encoder for learning the marginal image distributions of real images. Finally, pseudo patches with constrained noise vectors are put into RU-NET which is a component of our HybridGAN to generate a large number of synthetic MR images. In RU-NET, A SpliceLayer is then employed to fuse sub-images together in an interlaced manner into a full-size image. The experimental results show that HybridGAN can effectively synthesize a large variety of MR images and display a good visual quality. Compared to the state-of-the-art synthesis methods, our method achieves a significant improvement in terms of both visual and quantitative evaluation metrics.

Minimax Linear Estimation with the Probability Criterion under Unimodal Noise and Bounded Parameters

We consider a linear regression model with a vector of bounded parameters and a centered noise vector that has an uncertain unimodal distribution but known covariance matrix. We pose the minimax estimation problem for a linear combination of unknown parameters with the use of the probability criterion. The minimax estimate is determined as a result of minimizing a probability bound over the region of possible values of the variance and squared bias for all possible linear estimates. We establish that the resulting robust solution is less conservative in comparison with wider classes of distributions.

Private Information Retrieval Through Wiretap Channel II: Privacy Meets Security

We consider the problem of private information retrieval through wiretap channel II (PIR-WTC-II). In PIR-WTC-II, a user wants to retrieve a single message (file) privately out of M messages, which are stored in N replicated and non-communicating databases. An external eavesdropper observes a fraction μ <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">n</sub> (of its choice) of the traffic exchanged between the nth database and the user. In addition to the privacy constraint, the databases should encode the returned answer strings such that the eavesdropper learns absolutely nothing about the contents of the databases. We aim at characterizing the capacity of the PIR-WTC-II under the combined privacy and security constraints. We obtain a general upper bound for the problem in the form of a max-min optimization problem, which extends the converse proof of the PIR problem under asymmetric traffic constraints. We propose an achievability scheme that satisfies the security constraint by encoding a secret key, which is generated securely at each database, into an artificial noise vector using an MDS code. The user and the databases operate at one of the corner points of the achievable scheme for the PIR under asymmetric traffic constraints such that the retrieval rate is maximized under the imposed security constraint. The upper bound and the lower bound match for the case of M = 2 and M = 3 messages, for any N, and any μ = (μ <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">1</sub> , · · · , μ <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">N</sub> ).

Deep Reinforcement Learning-Based Adaptive Computation Offloading for MEC in Heterogeneous Vehicular Networks

The vehicular network needs efficient and reliable data communication technology to maintain low latency. It is very challenging to minimize the energy consumption and data communication delay while the vehicle is moving and wireless channels and bandwidth are time-varying. With the help of the emerging mobile edge computing (MEC) server, vehicles and roadside units (RSUs) can offload computing tasks to MEC associated with base station (BS). However, the environment for offloading tasks to MEC, e.g., wireless channel states and available bandwidth, is unstable. Therefore, ensuring the efficiency of computation offloading under such an unstable environment is a challenge. In this work, we design a task computation offloading model in a heterogeneous vehicular network; this model takes into account multiple stochastic tasks, the variety of wireless channels and bandwidth. To obtain the tradeoff between the cost of energy consumption and the cost of data transmission delay and avoid curse of dimensionality caused by the complexity of the large action space, we propose an adaptive computation offloading method based on deep reinforcement learning (ACORL) that can address the continuous action space. ACORL adds the Ornstein-Uhlenbeck (OU) noise vector to the action space with different factors for each action to validate the exploration. Multi transmission equipment can execute local processing and computation offloading to MEC. Nevertheless, ACORL considers the variety of wireless channels and available bandwidth between adjacent time slots. The numerical results illustrate that the proposed ACORL can effectively learn the optimal policy, which outperforms the Dueling DQN and greedy policy in the stochastic environment.

A novel DOA estimation algorithm using directional antennas in cylindrical conformal arrays

In this paper, a novel direction of arrival (DOA) estimation algorithm using directional antennas in cylindrical conformal arrays (CCAs) is proposed. To eliminate the shadow effect, we divide the CCAs into several subarrays to obtain the complete output vector. Considering the anisotropic radiation pattern of a CCA, which cannot be separated from the manifold matrix, an improved interpolation method is investigated to transform the directional subarray into omnidirectional virtual nested arrays without non-orthogonal perturbation on the noise vector. Then, the cross-correlation matrix (CCM) of the subarrays is used to generate the consecutive co-arrays without redundant elements and eliminate the noise vector. Finally, the full-rank equivalent covariance matrix is constructed using the output of co-arrays, and the unitary estimation of the signal parameters via rotational invariance techniques (ESPRIT) is performed on the equivalent covariance matrix to estimate the DOAs with low computational complexity. Numerical simulations verify the superior performance of the proposed algorithm, especially under a low signal-to-noise ratio (SNR) environment.

Generative adversarial networks with decoder–encoder output noises

In recent years, research on image generation has been developing very fast. The generative adversarial network (GAN) emerges as a promising framework, which uses adversarial training to improve the generative ability of its generator. However, since GAN and most of its variants use randomly sampled noises as the input of their generators, they have to learn a mapping function from a whole random distribution to the image manifold. As the structures of the random distribution and the image manifold are generally different, this results in GAN and its variants difficult to train and converge. In this paper, we propose a novel deep model called generative adversarial networks with decoder–encoder output noises (DE-GANs), which take advantage of both the adversarial training and the variational Bayesian inference to improve GAN and its variants on image generation performances. DE-GANs use a pre-trained decoder–encoder architecture to map the random noise vectors to informative ones and feed them to the generator of the adversarial networks. Since the decoder–encoder architecture is trained with the same data set as the generator, its output vectors, as the inputs of the generator, could carry the intrinsic distribution information of the training images, which greatly improves the learnability of the generator and the quality of the generated images. Extensive experiments demonstrate the effectiveness of the proposed model, DE-GANs.

Motion-Based Generator Model: Unsupervised Disentanglement of Appearance, Trackable and Intrackable Motions in Dynamic Patterns

Dynamic patterns are characterized by complex spatial and motion patterns. Understanding dynamic patterns requires a disentangled representational model that separates the factorial components. A commonly used model for dynamic patterns is the state space model, where the state evolves over time according to a transition model and the state generates the observed image frames according to an emission model. To model the motions explicitly, it is natural for the model to be based on the motions or the displacement fields of the pixels. Thus in the emission model, we let the hidden state generate the displacement field, which warps the trackable component in the previous image frame to generate the next frame while adding a simultaneously emitted residual image to account for the change that cannot be explained by the deformation. The warping of the previous image is about the trackable part of the change of image frame, while the residual image is about the intrackable part of the image. We use a maximum likelihood algorithm to learn the model parameters that iterates between inferring latent noise vectors that drive the transition model and updating the parameters given the inferred latent vectors. Meanwhile we adopt a regularization term to penalize the norms of the residual images to encourage the model to explain the change of image frames by trackable motion. Unlike existing methods on dynamic patterns, we learn our model in unsupervised setting without ground truth displacement fields or optical flows. In addition, our model defines a notion of intrackability by the separation of warped component and residual component in each image frame. We show that our method can synthesize realistic dynamic pattern, and disentangling appearance, trackable and intrackable motions. The learned models can be useful for motion transfer, and it is natural to adopt it to define and measure intrackability of a dynamic pattern.

The Estimation Performance of Nonlinear Least Squares for Phase Retrieval

Suppose that ${ { y}}= \lvert \text {A} { x}_{0}\rvert +\eta $ where ${ x}_{0}\in {\mathbb R} ^{{d}}$ is the target signal and $\eta \in {\mathbb R}^{{m}}$ is a noise vector. The aim of phase retrieval is to estimate ${x}_{0}$ from ${ { y}}$ . A popular model for estimating ${ x}_{0}$ is the nonlinear least squares ${\widehat { {x}}}:={\mathrm{ argmin}}_{ x} \| \lvert \text {A} { x}\rvert - { { y}}\|_{2}$ . One has already developed many efficient algorithms for solving the model, such as the seminal error reduction algorithm. In this paper, we present the estimation performance of the model with proving that $\| {\widehat { {x}}}- { x}_{0}\|\lesssim {\|\eta \|_{2}}/{\sqrt {{m}}}$ under the assumption of A being a Gaussian random matrix. We also prove the reconstruction error ${\|\eta \|_{2}}/{\sqrt {{m}}}$ is sharp. For the case where ${ x}_{0}$ is sparse, we study the estimation performance of both the nonlinear Lasso of phase retrieval and its unconstrained version. Our results are non-asymptotic, and we do not assume any distribution on the noise $\eta $ . To the best of our knowledge, our results represent the first theoretical guarantee for the nonlinear least squares and for the nonlinear Lasso of phase retrieval.

Reconstruction of shale image based on Wasserstein Generative Adversarial Networks with gradient penalty

Generative Adversarial Networks (GANs), as most popular artificial intelligence models in the current image generation field, have excellent image generation capabilities. Based on Wasserstein GANs with gradient penalty, this paper proposes a novel digital core reconstruction method. First, a convolutional neural network is used as a generative network to learn the distribution of real shale samples, and then a convolutional neural network is constructed as a discriminative network to distinguish reconstructed shale samples from real ones. Through this confrontation training method, realistic digital core samples of shale can be reconstructed. The paper uses two-point covariance function, Frechet Inception Distance and Kernel Inception Distance, to evaluate the quality of digital core samples of shale reconstructed by GANs. The results show that the covariance function can test the similarity between generated and real shale samples, and that GANs can efficiently reconstruct digital core samples of shale with high-quality. Compared with multiple point statistics, the new method does not require prior inference of the probability distribution of the training data, and directly uses noise vector to generate digital core samples of shale without using constraints of hard data in advance. It is easy to produce an unlimited number of new samples. Furthermore, the training time is also shorter, only 4 hours in this paper. Therefore, the new method has some good points compared with current methods. Cited as : Zha, W., Li, X., Xing, Y., He, L., Li, D. Reconstruction of shale image based on Wasserstein Generative Adversarial Networks with gradient penalty. Advances in Geo-Energy Research, 2020, 4(1): 107-114, doi: 10.26804/ager.2020.01.10

Mean-Square Performance of the Modified Filtered-x Affine Projection Algorithm

The modified filtered-x affine projection (MFxAP) algorithm is effective for active noise control owing to its good convergence behavior and medium computational burden. The transient and steady-state performances of the MFxAP algorithm have been analyzed in previous studies, which presented a relatively good agreement between the theory and measured results. However, the correlation between the weight-error vector and the past noise vectors is disregarded in the existing methods. Hence, a more accurate theoretical analysis for the MFxAP algorithm is presented herein, in which the effect of the past noise vector on the weight-error vector is considered comprehensively. Simulation results indicate that the proposed theoretical results match the experimental results more precisely than the previous studies, in particular, at the steady state.

Research on fault diagnosis and state assessment of vacuum pump based on acoustic emission sensors.

A vacuum pump is a widely used vacuum device and a key component of the space environment simulator. Aiming at the problem of fault diagnosis and state assessment of the vacuum pump, this paper proposes a complete set of empirical mode decomposition [Complete Ensemble Empirical Mode Decomposition with Adaptive Noise (CEEMDAN)] based on adaptive noise and support vector machine optimized by particle swarm optimization (PSO-SVM). The CEEMDAN method can adaptively decompose the acoustic emission signal of the vacuum pump to obtain several eigenmode functions [Intrinsic Mode Functions (IMFs)] and residuals. The normalized energy values of the IMF component are extracted as the eigenvector. The PSO algorithm is used to optimize the key parameters of the SVM, and the samples are used for training to establish a fault diagnosis model. The vacuum pump overload fault and vacuum pump with different working states are judged by experiments. The results show that the method has an accuracy of more than 97.0% and can effectively realize fault diagnosis and state assessment of vacuum pump equipment.