Search

Recent developments in the field of non-local attention (NLA) have led to a renewed interest in self-similarity-based single image super-resolution (SISR). Researchers usually use the NLA to explore non-local self-similarity (NSS) in SISR and achieve satisfactory reconstruction results. However, a surprising phenomenon that the reconstruction performance of the standard NLA is similar to that of the NLA with randomly selected regions prompted us to revisit NLA. In this paper, we first analyzed the attention map of the standard NLA from different perspectives and discovered that the resulting probability distribution always has full support for every local feature, which implies a statistical waste of assigning values to irrelevant non-local features, especially for SISR which needs to model long-range dependence with a large number of redundant non-local features. Based on these findings, we introduced a concise yet effective soft thresholding operation to obtain high-similarity-pass attention (HSPA), which is beneficial for generating a more compact and interpretable distribution. Furthermore, we derived some key properties of the soft thresholding operation that enable training our HSPA in an end-to-end manner. The HSPA can be integrated into existing deep SISR models as an efficient general building block. In addition, to demonstrate the effectiveness of the HSPA, we constructed a deep high-similarity-pass attention network (HSPAN) by integrating a few HSPAs in a simple backbone. Extensive experimental results demonstrate that HSPAN outperforms state-of-the-art approaches on both quantitative and qualitative evaluations. Our code and a pre-trained model were uploaded to GitHub (https://github.com/laoyangui/HSPAN) for validation. [ABSTRACT FROM AUTHOR]

We address the problem of person re-identification (reID), that is, retrieving person images from a large dataset, given a query image of the person of interest. A key challenge is to learn person representations robust to intra-class variations, as different persons could have the same attribute, and persons’ appearances look different, e.g., with viewpoint changes. Recent reID methods focus on learning person features discriminative only for a particular factor of variations (e.g., human pose), which also requires corresponding supervisory signals (e.g., pose annotations). To tackle this problem, we propose to factorize person images into identity-related and -unrelated features. Identity-related features contain information useful for specifying a particular person (e.g., clothing), while identity-unrelated ones hold other factors (e.g., human pose). To this end, we propose a new generative adversarial network, dubbed identity shuffle GAN (IS-GAN). It disentangles identity-related and -unrelated features from person images through an identity-shuffling technique that exploits identification labels alone without any auxiliary supervisory signals. We restrict the distribution of identity-unrelated features, or encourage the identity-related and -unrelated features to be uncorrelated, facilitating the disentanglement process. Experimental results validate the effectiveness of IS-GAN, showing state-of-the-art performance on standard reID benchmarks, including Market-1501, CUHK03 and DukeMTMC-reID. We further demonstrate the advantages of disentangling person representations on a long-term reID task, setting a new state of the art on a Celeb-reID dataset. Our code and models are available online: https://cvlab-yonsei.github.io/projects/ISGAN/. [ABSTRACT FROM AUTHOR]

Deep generative models have emerged as a powerful class of priors for signals in various inverse problems such as compressed sensing, phase retrieval and super-resolution. In this work, we consider the compressed sensing problem and assume the unknown signal to lie in the range of some pre-trained generative model. A popular approach for signal recovery is via gradient descent in the low-dimensional latent space. While gradient descent has achieved good empirical performance, its theoretical behavior is not well understood. We introduce the use of stochastic gradient Langevin dynamics (SGLD) for compressed sensing with a generative prior. Under mild assumptions on the generative model, we prove the convergence of SGLD to the true signal. We also demonstrate competitive empirical performance to standard gradient descent. [ABSTRACT FROM AUTHOR]

Contrastive self-supervised representation learning methods maximize the similarity between the positive pairs, and at the same time tend to minimize the similarity between the negative pairs. However, in general the interplay between the negative pairs is ignored as they do not put in place special mechanisms to treat negative pairs differently according to their specific differences and similarities. In this paper, we present Extended Momentum Contrast (XMoCo), a self-supervised representation learning method founded upon the legacy of the momentum-encoder unit proposed in the MoCo family configurations. To this end, we introduce a cross consistency regularization loss, with which we extend the transformation consistency to dissimilar images (negative pairs). Under the cross consistency regularization rule, we argue that semantic representations associated with any pair of images (positive or negative) should preserve their cross-similarity under pretext transformations. Moreover, we further regularize the training loss by enforcing a uniform distribution of similarity over the negative pairs across a batch. The proposed regularization can easily be added to existing self-supervised learning algorithms in a plug-and-play fashion. Empirically, we report a competitive performance on the standard Imagenet-1K linear head classification benchmark. In addition, by transferring the learned representations to common downstream tasks, we show that using XMoCo with the prevalently utilized augmentations can lead to improvements in the performance of such tasks. We hope the findings of this paper serve as a motivation for researchers to take into consideration the important interplay among the negative examples in self-supervised learning. [ABSTRACT FROM AUTHOR]

Optimization modeling of real-world application problems usually involves noise from various sources. Noisy optimization imposes challenges to optimization methods since the objective values can be different for multiple evaluations. In this article, we propose a novel online population size learning (OPL) technique of evolution strategies for handling noisy optimization problems. By re-evaluating a fraction of the candidates, we measure the strength of noise level of the re-evaluated candidate solutions and adapt the population size according to the noise level. The proposed OPL combines the advantages of both explicit averaging by re-evaluations and the implicit averaging by large population size and overcomes their limitations. We incorporate it with the covariance matrix adaptation evolution strategy (CMA-ES) and obtain OPL-CMA-ES. Compared with the existing noise handling technique, the proposed OPL is much simpler in both concepts and computation. We conduct comprehensive experiments to evaluate the algorithm’s performance on standard problems with Gaussian noise. We further evaluate the performance of OPL-CMA-ES on the black-box optimization benchmarks (BBOBs) noisy testbed, which is a standard platform for comparing black-box optimization algorithms, compared with the state-of-the-art noise-handling algorithms. The experimental results show that OPL-CMA-ES achieves remarkable performance and outperforms the compared variants. [ABSTRACT FROM AUTHOR]

The integration of phasor measurement units (PMUs) and phasor data concentrators (PDCs) in smart grids may be exploited by attackers to initiate new and sophisticated false data injection (FDI) attacks. Existing FDI attack mitigation approaches are generally less effective against sophisticated FDI attacks, such as collusive false data injection (CFDI) attacks launched by compromised PDCs (and PMUs) as we demonstrate in this paper. Thus, we propose a secure and resilience-enhanced scheme (SeCDM) to detect and mitigate such cyber threats in smart grids. Specifically, we design a decentralized homomorphic computation paradigm along with a hierarchical knowledge sharing algorithm to facilitate secure ciphertext calculation of state estimation residuals. Following this, a centralized FDI detector is implemented to detect FDI attacks. Findings from the security analysis demonstrate our approach achieves enhanced conventional FDI and CFDI attack resilience, and findings from our performance evaluations on the standard IEEE 14-, 24-, and 39-bus power systems also show that the communication overheads and computational complexity are reasonably “low”. [ABSTRACT FROM AUTHOR]

Existing person re-identification (Re-ID) methods usually rely heavily on large-scale thoroughly annotated training data. However, label noise is unavoidable due to inaccurate person detection results or annotation errors in real scenes. It is extremely challenging to learn a robust Re-ID model with label noise since each identity has very limited annotated training samples. To avoid fitting to the noisy labels, we propose to learn a prefatory model using a large learning rate at the early stage with a self-label refining strategy, in which the labels and network are jointly optimized. To further enhance the robustness, we introduce an online co-refining (CORE) framework with dynamic mutual learning, where networks and label predictions are online optimized collaboratively by distilling the knowledge from other peer networks. Moreover, it also reduces the negative impact of noisy labels using a favorable selective consistency strategy. CORE has two primary advantages: it is robust to different noise types and unknown noise ratios; it can be easily trained without much additional effort on the architecture design. Extensive experiments on Re-ID and image classification demonstrate that CORE outperforms its counterparts by a large margin under both practical and simulated noise settings. Notably, it also improves the state-of-the-art unsupervised Re-ID performance under standard settings. Code is available at https://github.com/mangye16/ReID-Label-Noise. [ABSTRACT FROM AUTHOR]

In this paper, we develop a novel visual-inertial navigation system with motion constraint (VINS-Motion), which extends the visual-inertial navigation system (VINS) to incorporate vehicle motion information for improving the autonomous vehicles localization accuracy. We introduce the vehicle stop detection module in the pre-processing measurement procedure. During the back-end processing, if the vehicle is detected to be in a stopping state, we perform optimization based on the stop constraint to help eliminate the abnormal jitter of the estimated pose, thus ensuring the reasonability of the trajectory. Otherwise, besides the prior information, IMU measurement residual, and visual measurement residual utilized in VINS, vehicle orientation/velocity constraint built by Ackerman steering model is first exploited to constitute residuals. We minimize the sum of priors and Mahalanobis norms of three kinds of residuals to obtain a maximum posterior estimation, thus increasing system consistency and accuracy. The proposed approach is validated on public datasets and compared with state-of-the-art algorithms, which demonstrates that the motion constraint proposed in this paper improves the standard VINS performance and achieves significantly higher positioning accuracy. [ABSTRACT FROM AUTHOR]

Computational surface modeling that underlies material recognition has transitioned from reflectance modeling using in-lab controlled radiometric measurements to image-based representations based on internet-mined single-view images captured in the scene. We take a middle-ground approach for material recognition that takes advantage of both rich radiometric cues and flexible image capture. A key concept is differential angular imaging, where small angular variations in image capture enables angular-gradient features for an enhanced appearance representation that improves recognition. We build a large-scale material database, Ground Terrain in Outdoor Scenes (GTOS) database, to support ground terrain recognition for applications such as autonomous driving and robot navigation. The database consists of over 30,000 images covering 40 classes of outdoor ground terrain under varying weather and lighting conditions. We develop a novel approach for material recognition called texture-encoded angular network (TEAN) that combines deep encoding pooling of RGB information and differential angular images for angular-gradient features to fully leverage this large dataset. With this novel network architecture, we extract characteristics of materials encoded in the angular and spatial gradients of their appearance. Our results show that TEAN achieves recognition performance that surpasses single view performance and standard (non-differential/large-angle sampling) multiview performance. [ABSTRACT FROM AUTHOR]

Three-phase pulsewidth-modulated converters require closed-loop controllers to meet the performance requirements and to achieve reliable responses over the entire operating range. Conventionally, by using the synchronous reference frame (SRF) representation, different linear and nonlinear control strategies have been proposed to improve the converter's dynamic performance. However, the achieved improvements cannot be objectively assessed due to the absence of a performance limit standard reference of the converter's dynamics. In this article, the theoretical minimum time solutions for three-phase converters in the SRF are derived and analyzed, providing an objective reference point for the system's dynamic performance evaluation. The normalized dynamic model of the converter in the SRF is derived and combined with the time-optimal control principle to find the theoretical minimum transient time solutions. The state-plane representation of these minimum time responses enables a straightforward and comprehensive interpretation of the converter's dynamic behavior, and it simplifies the derivation of closed-form expressions that characterize the minimum time transient parameters. The introduced analysis, minimum time dynamic responses, derived transient's expressions, and dynamic performance evaluation are validated by simulation and experimental results. [ABSTRACT FROM AUTHOR]

High-rate product codes (PCs) and staircase codes (SCs) are ubiquitous codes in high-speed optical communication achieving near-capacity performance on the binary symmetric channel. Their success is mostly due to very efficient iterative decoding algorithms that require very little complexity. In this paper, we extend the density evolution (DE) analysis for PCs and SCs to a channel with ternary output and ternary message passing, where the third symbol marks an erasure. We investigate the performance of a standard error-and-erasure decoder and of its simplification using DE. The proposed analysis can be used to find component code configurations and quantizer levels for the channel output. We also show how the use of even-weight BCH subcodes as component codes can improve the decoding performance at high rates. The DE results are verified by Monte-Carlo simulations, which show that additional coding gains of up to 0.6dB are possible by ternary decoding, at only a small additional increase in complexity compared to traditional binary message passing. [ABSTRACT FROM AUTHOR]

In order for a closed-loop radar system to adapt properly to changing operational conditions, the performance standards and requirements of the radar system must be known. Often, these performance standards are contained in statements made in language, which describe the different goals of the radar system. In this article, we explore the use of language-based cost functions (LBCFs) to encode performance standards made in imprecise language into an objective function, which a fully adaptive radar (FAR) can use to optimize its performance in real time. To enable the LBCFs, we also propose a statement decomposer, which takes in statements as inputs and yields the information required by LBCFs as outputs. A simulation study of a FAR tracking a running human is conducted with four different cost function methodologies: quadratic, weighted global criterion, fuzzy optimization, and LBCFs. Results show that when statements are properly formulated and presented, the statement decomposer is able to successfully extract the relevant information. Results also show that when paired with a previously developed optimization routine, the LBCF system is capable of autonomously creating cost functions that outperform the other cost functions in the simulated FAR target tracking scenario. [ABSTRACT FROM AUTHOR]

5G is envisioned to simultaneously provide diverse service types with heterogeneous needs under very different application scenarios and business models. Therefore, network slicing is included as a key feature of the 5G architecture to allow sharing a common infrastructure among different tenants, such as mobile communication providers, vertical market players, etc. In order to provide the Radio Access Network (RAN) with network slicing capabilities, mechanisms that efficiently distribute the available capacity among the different tenants while satisfying their needs are required. For this purpose, this paper proposes a multi-agent reinforcement learning approach for RAN capacity sharing. It makes use of the Deep Q-Network algorithm in a way that each agent is associated to a different tenant and learns the capacity to be provided to this tenant in each cell while ensuring that the service level agreements are satisfied and that the available radio resources are efficiently used. The consideration of multiple agents contributes to a better scalability and higher learning speed in comparison to single-agent approaches. In this respect, results show that the policy learnt by the agent of one tenant can be generalised and directly applied by other agents, thus reducing the complexity of the training and making the proposed solution easily scalable, e.g., to add new tenants in the system. The proposed approach is well aligned with the on-going 3GPP standardization work and guidelines for the parametrization of the solution are provided, thus enforcing its practical applicability. [ABSTRACT FROM AUTHOR]

We consider the problem of conducting frame rate dependent video quality assessment (VQA) on videos of diverse frame rates, including high frame rate (HFR) videos. More generally, we study how perceptual quality is affected by frame rate, and how frame rate and compression combine to affect perceived quality. We devise an objective VQA model called Space-Time GeneRalized Entropic Difference (GREED) which analyzes the statistics of spatial and temporal band-pass video coefficients. A generalized Gaussian distribution (GGD) is used to model band-pass responses, while entropy variations between reference and distorted videos under the GGD model are used to capture video quality variations arising from frame rate changes. The entropic differences are calculated across multiple temporal and spatial subbands, and merged using a learned regressor. We show through extensive experiments that GREED achieves state-of-the-art performance on the LIVE-YT-HFR Database when compared with existing VQA models. The features used in GREED are highly generalizable and obtain competitive performance even on standard, non-HFR VQA databases. The implementation of GREED has been made available online: https://github.com/pavancm/GREED. [ABSTRACT FROM AUTHOR]

This article presents a comparison of measurement methods for current and voltage distortions in low-voltage networks in the frequency range from 2 to 150 kHz (supraharmonics). The comparison encompasses the methods informatively described in IEC and CISPR international standards, as well as other innovative techniques presented in the literature. This work is carried out within a novel framework that includes advanced and complex synthetic test signals, as well as real grid recordings that allow an accurate comparison of the performance of the tested methods. Specifically designed indices are employed to characterize the accuracy of the tested methods in the frequency and amplitude assessments. In light of that, the strengths and weaknesses of the methods are identified. The results of this article contribute to the ongoing standardization work carried out by the IEC SC77A/WG9 with the purpose of defining a normative measurement method suitable for assessing grid disturbance levels in the range from 2 to 150 kHz. [ABSTRACT FROM AUTHOR]

This article proposes an online scheme for state estimation of a generic class of nonlinear dynamical systems in the presence of abnormal measurement data from sensors. We thoroughly illustrate why the performance of standard recursive Bayesian inference degrades in the presence of any measurement distortion. After demonstrating how different abnormalities can be accommodated explicitly inside a generic state–space model, we propose a robust mechanism to perform recursive Bayesian inference on the presented model to not only detect but also mitigate the effect of corrupted measurements in the final state estimates. Using simulations and experimentation, we demonstrate the success of the proposed framework in reducing the impact of different types of distortions in measurements. The ability to tackle different kinds of measurement abnormalities during online inference sets the proposed method apart from the existing techniques. [ABSTRACT FROM AUTHOR]

The article reports that Standards are fundamental in creating and maintaining an open and competitive market for equipment manufacturers and in assuring national and international interoperability of telecommunications technology and processes. Our interactive world is aided and achieved by standards which may be based on patented technologies that enable all types of devices to communicate efficiently with each other.