Multiscale Strategy Research Articles

Multiscale simulation of hygrothermal stress and mechanical properties degradation of carbon fiber reinforced polymers laminates

AbstractA mesoscopic representative volume element (RVE) model of carbon fiber reinforced polymers (CFRP) laminate with 15 plies was established based on the principle of fiber random disturbance. The RVE model was validated by comparing the elastic parameters calculated by the RVE model, Chamis model, and Bridge model. Wet stress and thermal stress were coupled using a sequential method. Macroscopic and microscopic subroutines were developed to calculate the stress–strain field of CFRP laminate and RVE model. A multiscale numerical simulation of three‐point bending of moisture‐saturated CFRP laminate was conducted. Failure process and failure modes of CFRP laminate were discussed based on the multiscale simulation and SEM observations. The results show that the RVE model and the multiscale simulation strategy presented in this work can predict the coupled hygrothermal stress and the strength degradation due to moisture absorption very well. The maximum error between the simulated and experimental averaged force of the moisture‐saturated CFRP laminate subjected to three‐point bending is about 3.4%. Multiscale simulation demonstrated that uneven distribution of fibers can result in the fiber/matrix interfaces bearing a larger transverse stress in the fiber's dense distribution area. It is the main reason that results in fiber/matrix interface failure of CFRP laminates working in hygrothermal environments.Highlights A RVE model of CFRP laminate was presented using fibers random disturbance. Coupled hygrothermal stress of single fiber and surrounded matrix was obtained. A multiscale simulation of the strength change of the CFRP plate due to wet‐thermal stress was conducted. Failure modes of CFRP laminate are discussed based on the multiscale simulation.

High‐Remanence Fe‐Rich 2:17‐Type Sm–Co Sintered Magnets Mediated via Multiscale Microstructure

Realizing high remanence in Fe‐rich 2:17‐type Sm–Co sintered magnets remains a challenge when Fe content exceeds 23 wt%. Herein, a record remanence of 12.42 kGs is obtained by multiscale microstructure regulation strategy, based on co‐optimization from pulverizing treatment to solid solution and aging treatment. The density of sintered magnets is significantly enhanced by refining magnetic powders using ball milling to reduce porosity. Grain boundary precipitation phase is eliminated via prolonging solid solution time, together with reduced Fe concentration in the cell boundaries, due to increased proportion of defected 1:3R planar Z phase. Furthermore, the proportion of 1:3R planar Z phase is increased by greatly reducing the 2:17R’ phase during appropriately isothermal aging temperature to reduce Fe concentration in cell boundaries and the remanence of the magnet increases. The studies deepen the understanding of the correlation among process, microstructure, and remanence, guiding the preparation of Fe‐rich 2:17‐type Sm–Co sintered magnets with high magnetic performance.

Inter-Domain Invariant Cross-Domain Object Detection Using Style and Content Disentanglement for In-Vehicle Images

The accurate detection of relevant vehicles, pedestrians, and other targets on the road plays a crucial role in ensuring the safety of autonomous driving. In recent years, object detectors based on Transformers or CNNs have achieved excellent performance in the fully supervised paradigm. However, when the trained model is directly applied to unfamiliar scenes where the training data and testing data have different distributions statistically, the model’s performance may decrease dramatically. To address this issue, unsupervised domain adaptive object detection methods have been proposed. However, these methods often exhibit decreasing performance when the gap between the source and target domains increases. Previous works mainly focused on utilizing the style gap to reduce the domain gap while ignoring the content gap. To tackle this challenge, we introduce a novel method called IDI-SCD that effectively addresses both the style and content gaps simultaneously. Firstly, the domain gap is reduced by disentangling it into the style gap and content gap, generating corresponding intermediate domains in the meanwhile. Secondly, during training, we focus on one single domain gap at a time to achieve inter-domain invariance. That is, the content gap is tackled while maintaining the style gap, and vice versa. In addition, the style-invariant loss is used to narrow down the style gap, and the mean teacher self-training framework is used to narrow down the content gap. Finally, we introduce a multiscale fusion strategy to enhance the quality of pseudo-labels, which mainly focus on enhancing the detection performance for extreme-scale objects (very large or very small objects). We conduct extensive experiments on four mainstream datasets of in-vehicle images. The experimental results demonstrate the effectiveness of our method and its superiority over most of the existing methods.

Learn from orientation prior for radiograph super-resolution: Orientation operator transformer

Background and objective: High-resolution radiographic images play a pivotal role in the early diagnosis and treatment of skeletal muscle-related diseases. It is promising to enhance image quality by introducing single-image super-resolution (SISR) model into the radiology image field. However, the conventional image pipeline, which can learn a mixed mapping between SR and denoising from the color space and inter-pixel patterns, poses a particular challenge for radiographic images with limited pattern features. To address this issue, this paper introduces a novel approach: Orientation Operator Transformer - O2former. Methods: We incorporate an orientation operator in the encoder to enhance sensitivity to denoising mapping and to integrate orientation prior. Furthermore, we propose a multi-scale feature fusion strategy to amalgamate features captured by different receptive fields with the directional prior, thereby providing a more effective latent representation for the decoder. Based on these innovative components, we propose a transformer-based SISR model, i.e., O2former, specifically designed for radiographic images. Results: The experimental results demonstrate that our method achieves the best or second-best performance in the objective metrics compared with the competitors at ×4 upsampling factor. For qualitative, more objective details are observed to be recovered. Conclusions: In this study, we propose a novel framework called O2former for radiological image super-resolution tasks, which improves the reconstruction model's performance by introducing an orientation operator and multi-scale feature fusion strategy. Our approach is promising to further promote the radiographic image enhancement field.

Semi-Supervised Object Detection with Multi-Scale Regularization and Bounding Box Re-Prediction

Semi-supervised object detection has become a hot topic in recent years, but there are still some challenges regarding false detection, duplicate detection, and inaccurate localization. This paper presents a semi-supervised object detection method with multi-scale regularization and bounding box re-prediction. Specifically, to improve the generalization of the two-stage object detector and to make consistent predictions related to the image and its down-sampled counterpart, a novel multi-scale regularization loss is proposed for the region proposal network and the region-of-interest head. Then, in addition to using the classification probabilities of the pseudo-labels to exploit the unlabeled data, this paper proposes a novel bounding box re-prediction strategy to re-predict the bounding boxes of the pseudo-labels in the unlabeled images and select the pseudo-labels with reliable bounding boxes (location coordinates) to improve the model’s localization accuracy based on its unsupervised localization loss. Experiments on the public MS COCO and Pascal VOC show that our proposed method achieves a competitive detection performance compared to other state-of-the-art methods. Furthermore, our method offers a multi-scale regularization strategy and a reliably located pseudo-label screening strategy, both of which facilitate the development of semi-supervised object detection techniques and boost the object detection performance in autonomous driving, industrial inspection, and agriculture automation.

Well-balanced performance achieved in PZT piezoceramics via a multiscale regulation strategy.

Emerging high-power piezoelectric applications demand the development of piezoelectric materials featuring both a high mechanical quality factor (Qm) and a large piezoelectric coefficient (d33). However, it is widely accepted that an increase in d33 is usually accompanied with a decrease in Qm, and vice versa. Herein, a multiscale regulation strategy is proposed to improve Qm and d33 simultaneously from the perspectives of phase structure, ferroelectric domains, and lattice defects. A well-balanced combination of electromechanical performances with Qm = 726, d33 = 502 pC N-1, kp = 0.69, tan δ = 0.0024, and TC = 267 °C was obtained. Through structural characterization, it was observed that the morphotropic phase boundary and enhanced dispersion behavior lead to a lowered energy barrier, which contributes to polarization rotation and enhances piezoelectric performance. At the same time, the excellent piezoelectric performances also benefit from the highly oriented domain structure and small domain size after high-temperature poling. Furthermore, the segregation of Ba2+ causes A-site defects in the crystal lattice, accompanied with an increase in oxygen vacancies, which maintains the hardening effect of the ceramics. This study proposes a multiscale regulation strategy, providing insights for the design of high-power piezoelectric ceramics with high d33 and Qm.

MSFlow: Multiscale Flow-Based Framework for Unsupervised Anomaly Detection.

Unsupervised anomaly detection (UAD) attracts a lot of research interest and drives widespread applications, where only anomaly-free samples are available for training. Some UAD applications intend to locate the anomalous regions further even without any anomaly information. Although the absence of anomalous samples and annotations deteriorates the UAD performance, an inconspicuous, yet powerful statistics model, the normalizing flows, is appropriate for anomaly detection (AD) and localization in an unsupervised fashion. The flow-based probabilistic models, only trained on anomaly-free data, can efficiently distinguish unpredictable anomalies by assigning them much lower likelihoods than normal data. Nevertheless, the size variation of unpredictable anomalies introduces another inconvenience to the flow-based methods for high-precision AD and localization. To generalize the anomaly size variation, we propose a novel multiscale flow-based framework (MSFlow) composed of asymmetrical parallel flows followed by a fusion flow to exchange multiscale perceptions. Moreover, different multiscale aggregation strategies are adopted for image-wise AD and pixel-wise anomaly localization according to the discrepancy between them. The proposed MSFlow is evaluated on three AD datasets, significantly outperforming existing methods. Notably, on the challenging MVTec AD benchmark, our MSFlow achieves a new state-of-the-art (SOTA) with a detection AUORC score of up to 99.7%, localization AUCROC score of 98.8% and PRO score of 97.1%.

Deep Attentional Guided Image Filtering.

Guided filter is a fundamental tool in computer vision and computer graphics, which aims to transfer structure information from the guide image to the target image. Most existing methods construct filter kernels from the guidance itself without considering the mutual dependency between the guidance and the target. However, since there typically exist significantly different edges in two images, simply transferring all structural information from the guide to the target would result in various artifacts. To cope with this problem, we propose an effective framework named deep attentional guided image filtering, the filtering process of which can fully integrate the complementary information contained in both images. Specifically, we propose an attentional kernel learning module to generate dual sets of filter kernels from the guidance and the target and then adaptively combine them by modeling the pixelwise dependency between the two images. Meanwhile, we propose a multiscale guided image filtering module to progressively generate the filtering result with the constructed kernels in a coarse-to-fine manner. Correspondingly, a multiscale fusion strategy is introduced to reuse the intermediate results in the coarse-to-fine process. Extensive experiments show that the proposed framework compares favorably with the state-of-the-art methods in a wide range of guided image filtering applications, such as guided super-resolution (SR), cross-modality restoration, and semantic segmentation. Moreover, our scheme achieved the first place in the real depth map SR challenge held in ACM ICMR 2021. The codes can be found at https://github.com/zhwzhong/DAGF.

Flexible Iris Matching Based on Spatial Feature Reconstruction

In an iris recognition-at-a-distance (IAAD) system used in surveillance scenarios, the camera usually captures a large number of low-quality images. These images exhibit partial occlusions due to eyelids and eyelashes, specular reflections, and severe deformations caused by pupil dilations and contractions. Recognizing these low-quality images is a challenging yet dominant problem in IAAD. To mitigate this issue, current iris recognition systems mostly filter out low-quality images by using strict criteria based on image quality evaluation. This strategy, however, wastes device capabilities and produces low throughput of subjects. Other systems require highly cooperative users. In this work, we propose a novel occlusion-robust, deformation-robust, and alignment-free framework for low-quality iris matching, which integrates the merits of deep features and sparse representation in an end-to-end learning process known as iris spatial feature reconstruction (ISFR). Here each probe image can be sparsely reconstructed on the basis of appropriate feature maps from gallery high-quality images. ISFR uses the error from robust reconstruction over spatial pyramid features to measure similarities between two iris images, which naturally avoids the time-consuming alignment step. In summary, the distinctiveness of deep features, the robustness of sparse reconstruction, and the flexibility of multiscale matching strategy are unified in a general framework to attain more accurate and reasonable iris matching. Extensive experimental results on four public iris image databases demonstrate that the proposed method significantly outperforms both traditional and deep learning-based iris recognition methods.

A CNN-Based Multi-Scale Pooling Strategy for Acoustic Scene Classification

A CNN-Based Multi-Scale Pooling Strategy for Acoustic Scene Classification

Image deblurring method based on self-attention and residual wavelet transform

The restoration technology of non-uniform blurred images is a challenging open topic. Most of the existing algorithms fail to effectively fuse multi-scale feature extraction with a self-attention mechanism, and also ignore the potential contribution of image frequency domain information to image restoration. Frequency domain features play an important role in restoring high-quality images and ignoring this property often leads to over-smoothing of the restoration results. In response to these problems, an image deblurring method based on self-attention and residual wavelet transform is proposed in this paper. Based on a single U-Net network, the multi-scale feature cross-fusion strategy and self-attention mechanism are combined to make the network pay more attention to different degrees of blurred regions so that relatively robust blurred features can be extracted for image deblurring. Meanwhile, considering the important role of frequency domain features for image restoration, the wavelet transform is embedded into the depth residual network to convert spatial domain features to wavelet domain, and the sharp details such as edge contours of blurred features are restored by making full use of the texture structure information possessed by high-frequency sub-bands, which further improves the image restoration performance. Experimental results of quantitative and qualitative comparison with other state-of-the-art methods show that the image deblurring effect of the proposed network performs favorably in terms of peak signal-to-noise ratio (PSNR) and structural similarity (SSIM) metrics. The codes and models are available at https://github.com/BingY998/MRDNet.

MEConformer: Highly representative embedding extractor for speaker verification via incorporating selective convolution into deep speaker encoder

Transformer models have demonstrated superior performance across various domains, including computer vision, natural language processing, and speech recognition. The success of these models can be attributed to their robust parallel capacity and high computation speed, primarily reliant on the attention layer. In the domain of speaker recognition, state-of-the-art results have been achieved using convolutional neural network (CNN) architectures, particularly with speaker embeddings represented by x-vectors and r-vectors. However, existing CNN-based methods tend to focus on local features while overlooking the global dependence of voiceprint features, resulting in the loss of crucial information. Moreover, the presence of noise in audio data is an influential factor that cannot be disregarded, as it significantly impacts the extraction of discriminative speaker embeddings. To address these challenges, we propose a novel model called the Multi-Scale Expand Convolution Transformer (MEConformer). This model aims to convert variable-length audio into a fixed low-dimensional representation. The MEConformer leverages a CNN framework with expanded receptive fields to capture frame-level features effectively. Additionally, we introduce a transformer encoder that incorporates contextual dependencies, enabling the extraction of both frame-level and discourse-level feature representations. Furthermore, we present a multi-scale residual aggregation strategy, which facilitates the efficient transmission of voiceprint information across the model. By combining these innovative components, the MEConformer achieves a state-of-the-art Equal Error Rate (EER) of 3.72% on the VoxCeleb1 test set. Furthermore, it demonstrates EERs of 5.94% and 3.72% on the VoxCeleb1-H and VoxCeleb-E datasets, respectively. The code for the proposed MEConformer model will be made publicly available at https://codeocean.com/capsule/4563012/tree.

Protecting Rural Large Old Trees with Multi-Scale Strategies: Integrating Spatial Analysis and the Contingent Valuation Method (CVM) for Socio-Cultural Value Assessment

Governments are faced with the unique challenge of implementing large-scale and targeted protection against the global decline of large old trees. Incorporating socio-cultural values and encouraging public involvement are important parts of conservation policy. However, current studies on the socio-cultural valuation of large old trees are still limited, and how rural residents perceive the human-related value of large old trees remains largely unknown. Using a quantitative, spatial analysis and the contingent valuation method (CVM), we tried to explore a multi-scale socio-cultural valuation and protection framework based on a case study of Baoding City and Xiongan New Area in North China. The results showed that (1) the scattered large old trees in the study area were generally at a relatively younger stage, showing normal growth performance but having poor living environments. Some 96.99% of the trees resided in the countryside. Their distribution showed an agglomerative pattern with several clusters. (2) The species richness was relatively lower than that reported in urban areas. The species diversity had an obvious high–low gradient from the mountain to plain areas. Most endemic species were found in habitats of the village fringe (VF) and government/community/institutional ground (GC). (3) The mean willingness to pay (WTP) for the socio-cultural value of scattered large old trees was CNY 132.48 per year per person (1 US dollar equals about 7.2 CNY) of all the respondents, and CNY 84.30 per year per person with regard to farmers, which is relatively higher than that reported in large cities. (4) Economic income, gender, age, education level, place of residence, diameter at breast height, and tree habitat were factors that significantly influenced the WTP, among which economic income was the most significant. (5) The importance ranking of socio-cultural value connotations perceived by rural residents was as follows: spiritual attachment and homesickness > fengshui > social bond > witnessing history > education > creative inspiration. (6) The annual gross value was estimated to be CNY 349 million in the study area, and CNY 169,500 for a single tree on average. Based on the case study, a conceptual framework for socio-cultural value assessment and multi-scale protection of large old trees was proposed, which can provide references for the improvement of current conservation policies from both quantitative and qualitative perspectives, and give insights into rural revitalization strategies in China.

Multiscale Structural Design of 2D Nanomaterials-based Flexible Electrodes for Wearable Energy Storage Applications.

2D nanomaterials play a critical role in realizing high-performance flexible electrodes for wearable energy storge devices, owing to their merits of large surface area, high conductivity and high strength. The electrode is a complex system and the performance is determined by multiple and interrelated factors including the intrinsic properties of materials and the structures at different scales from macroscale to atomic scale. Multiscale design strategies have been developed to engineer the structures to exploit full potential and mitigate drawbacks of 2D materials. Analyzing the design strategies and understanding the working mechanisms are essential to facilitate the integration and harvest the synergistic effects. This review summarizes the multiscale design strategies from macroscale down to micro/nano-scale structures and atomic-scale structures for developing 2D nanomaterials-based flexible electrodes. It starts with brief introduction of 2D nanomaterials, followed by analysis of structural design strategies at different scales focusing on the elucidation of structure-property relationship, and ends with the presentation of challenges and future prospects. This review highlights the importance of integrating multiscale design strategies. Finding from this review may deepen the understanding of electrode performance and provide valuable guidelines for designing 2D nanomaterials-based flexible electrodes.

Electrophysiological brain imaging based on simulation-driven deep learning in the context of epilepsy

Identifying the location, the spatial extent and the electrical activity of distributed brain sources in the context of epilepsy through ElectroEncephaloGraphy (EEG) recordings is a challenging task because of the highly ill-posed nature of the underlying Electrophysiological Source Imaging (ESI) problem. To guarantee a unique solution, most existing ESI methods pay more attention to solve this inverse problem by imposing physiological constraints. This paper proposes an efficient ESI approach based on simulation-driven deep learning. Epileptic High-resolution 256-channels scalp EEG (Hr-EEG) signals are simulated in a realistic manner to train the proposed patient-specific model. More particularly, a computational neural mass model developed in our team is used to generate the temporal dynamics of the activity of each dipole while the forward problem is solved using a patient-specific three-shell realistic head model and the boundary element method. A Temporal Convolutional Network (TCN) is considered in the proposed model to capture local spatial patterns. To enable the model to observe the EEG signals from different scale levels, the multi-scale strategy is leveraged to capture the overall features and fine-grain features by adjusting the convolutional kernel size. Then, the Long Short-Term Memory (LSTM) is used to extract temporal dependencies among the computed spatial features. The performance of the proposed method is evaluated through three different scenarios of realistic synthetic interictal Hr-EEG data as well as on real interictal Hr-EEG data acquired in three patients with drug-resistant partial epilepsy, during their presurgical evaluation. A performance comparison study is also conducted with two other deep learning-based methods and four classical ESI techniques. The proposed model achieved a Dipole Localization Error (DLE) of 1.39 and Normalized Hamming Distance (NHD) of 0.28 in the case of one patch with SNR of 10 dB. In the case of two uncorrelated patches with an SNR of 10 dB, obtained DLE and NHD were respectively 1.50 and 0.28. Even in the more challenging scenario of two correlated patches with an SNR of 10 dB, the proposed approach still achieved a DLE of 3.74 and an NHD of 0.43. The results obtained on simulated data demonstrate that the proposed method outperforms the existing methods for different signal-to-noise and source configurations. The good behavior of the proposed method is also confirmed on real interictal EEG data. The robustness with respect to noise makes it a promising and alternative tool to localize epileptic brain areas and to reconstruct their electrical activities from EEG signals.

Multiscale Synergetic Bandgap/Structure Engineering in Semiconductor Nanofibrous Aerogels for Enhanced Solar Evaporation.

Solar-driven interface evaporation has been identified as a sustainable seawater desalination and water purification technology. Nonetheless, the evaporation performance is still restricted by salt deposition and heat loss owing to weak solar spectrum absorption, tortuous channels, and limited plane area of conventional photothermal material. Herein, the semiconductor nanofibrous aerogels with a narrow bandgap, vertically aligned channels, and a conical architecture are constructed by the multiscale synergetic engineering strategy, encompassing bandgap engineering at the atomic scale and structure engineering at the nano-micro scale. As a proof-of-concept demonstration, a Co-doped MoS2 nanofibrous aerogel is synthesized, which exhibits the entire solar absorption, superhydrophilic, and excellent thermal insulation, achieving a net evaporation rate of 1.62 kg m-2 h-1 under 1 sun irradiation, as well as a synergistically efficient dye ion adsorption function. This work opens up new possibilities for the development of solar evaporators for practical applications in clean water production.

Multi-scale style generative and adversarial contrastive networks for single domain generalization fault diagnosis

Domain generalization methods can effectively identify machinery faults under unseen new target working conditions. Nevertheless, most of them rely on data from multiple source domains that are available for model training. However, it is laborious difficult to collect complete monitoring data of machinery under multiple working conditions. Confronting the scenario that only one working condition is available, this paper proposes a novel single domain generalization model, termed multi-scale style generative and adversarial contrastive networks (MSG-ACN), which learns diagnosis knowledge from the single working condition and generalizes it to new working conditions. The main idea of the MSG-ACN model is to generate diverse samples in an extended domain via a domain generation module, and extract domain-invariant features from the source and extended domains via a diagnosis task module. A multi-scale style generation strategy is established to ensure that the generated samples contain abundant state information with the aids of multi-scale convolutional kernels and style learning. Furthermore, an adversarial contrastive learning strategy is designed to promote the learning of class-wise domain-invariant representations while maintaining the diversity of the generated samples. Extensive generalization diagnosis experiments on two datasets verify the superiority of the proposed method over the state-of-the-art fault diagnosis methods.

Multi-scale Coefficients Fusion Strategy for Enhancement of SAM Image in Solder Joints Detection

Multi-scale Coefficients Fusion Strategy for Enhancement of SAM Image in Solder Joints Detection

Tough and Moldable Sustainable Cellulose-Based Structural Materials via Multiscale Interface Engineering.

All-natural materials derived from cellulose nanofibers (CNFs) are expected to be used to replace engineering plastics and have attracted much attention. However, the lack of crack extension resistance and 3D formability of nanofiber-based structural materials hinders their practical applications. Here, a multiscale interface engineering strategy is reported to construct high-performance cellulose-based materials. The sisal microfibers are surface treated to expose abundant active CNFs with positive charges, thereby enhancing their interfacial combination with the negatively charged CNFs. The robust multiscale dual network enables easy molding of multiscale cellulose-based structural materials into complex 3D special-shaped structures, resulting in nearly twofold and fivefold improvements in toughness and impact resistance compared with those of CNFs-based materials. Moreover, this multiscale interface engineering strategy endows cellulose-based structural materials with better comprehensive performance than petrochemical-based plastics and broadens cellulose's potential for lightweight applications as structural materials with lower environmental effects.

Visible light sensing based on shadow features using multi-scale region convolutional neural network.

There are various production items in the industrial internet of things (IIoT) environment, such as pedestrians, robots, automated automated guided vehicles, etc. The practice industrial environment requires simultaneous communication and sensing of production items to achieve intelligent production and control. Thus, sensing methods not only require the integration of communication but also achieve sensing tasks such as recognition and positioning. Compared with traditional sensing media, visible light sensing has the advantages of high-speed communication, high sensing accuracy, and security, low energy consumption, and has become a potential sensing technology. Based on the strong directivity of visible light spatial radiation and the consistency of light intensity and position, this paper proposes a multi-scale visible light sensing-region convolutional neural network (VLS-RCNN) framework based on shadow features for multiple target sensing. The framework enables the recognition and positioning to use shared visible light shadow features to assist each other, and the multi-scale compensation strategy of the shadow region makes the framework more robust. The simulation results show that positioning results in the sensing area improve the recognition accuracy. The recognition results also reduce the positioning error without additional overhead. Therefore, this paper provides a new perspective for the sensing technology in the future IIoT, which should be considered to sense objects of interest by utilizing the inherent characteristics of visible light.