Rich Detail Research Articles

A multi-stage, first-order phase transition in LaFe11.8Si1.2: Interplay between the structural, magnetic, and electronic degrees of freedom

Alloys with a first-order magnetic transition are central to solid-state refrigeration technology, sensors and actuators, or spintronic devices. The discontinuous nature of the transition in these materials is a consequence of the coupling between the magnetic, electronic, and structural subsystems, and such transition can, in principle, cross several metastable states, where at one point, the transition takes place within the magnetic subsystem, while at another, the changes occur in the structural or electronic subsystems. To address this issue, we conducted simultaneous measurements of the macroscopic properties—magnetization, temperature change of the sample, longitudinal, and transversal magnetostrictions—to reveal the rich details of the magneto-structural, first-order transition occurring in the prototypical alloy LaFe11.8Si1.2. We found that the transition does not complete in one but in two distinct stages. The presence of the intermediate state changes the potential-energy landscape, which then impacts strongly on the width of the hysteresis associated with the first-order transition. We complement these findings with experiments on the atomistic scale, i.e., x-ray absorption spectroscopy, x-ray magnetic circular dichroism, and Mössbauer spectroscopy, and then combine them with first-principles calculations to reveal the full complexity and two-stage nature of the transition. This new approach can be successfully extended to a large class of advanced magnetic materials that exhibit analogous transformations.

Pyramid-VAE-GAN: Transferring hierarchical latent variables for image inpainting

Significant progress has been made in image inpainting methods in recent years. However, they are incapable of producing inpainting results with reasonable structures, rich detail, and sharpness at the same time. In this paper, we propose the Pyramid-VAE-GAN network for image inpainting to address this limitation. Our network is built on a variational autoencoder (VAE) backbone that encodes high-level latent variables to represent complicated high-dimensional prior distributions of images. The prior assists in reconstructing reasonable structures when inpainting. We also adopt a pyramid structure in our model to maintain rich detail in low-level latent variables. To avoid the usual incompatibility of requiring both reasonable structures and rich detail, we propose a novel cross-layer latent variable transfer module. This transfers information about long-range structures contained in high-level latent variables to low-level latent variables representing more detailed information. We further use adversarial training to select the most reasonable results and to improve the sharpness of the images. Extensive experimental results on multiple datasets demonstrate the superiority of our method. Our code is available at https://github.com/thy960112/Pyramid-VAE-GAN.

Adaptive Total-Variation and Nonconvex Low-Rank Model for Image Denoising

In recent years, image denoising methods based on total variational regularization have attracted extensive attention. However, the traditional total variational regularization method is an approximate solution based on convex method, and does not consider the particularity of the region with rich details. In this paper, the adaptive total-variation and nonconvex low-rank model for image denoising is proposed, which is a hybrid regularization model. First, the image is decomposed into sparse terms and low rank terms, and then the total variational regularization is used to denoise. At the same time, an adaptive coefficient based on gradient is constructed to adaptively judge the flat area and detail texture area, slow down the denoising intensity of detail area, and then play the role of preserving detail information. Finally, by constructing a nonconvex function, the optimal solution of the function is obtained by using the alternating minimization method. This method not only effectively removes the image noise, but also retains the detailed information of the image. The experimental results show the effectiveness of the proposed model, and SNR and SSIM of the denoised image are improved.

Image inpainting via progressive decoder and gradient guidance

Very recently, with the widespread research of deep learning, its achievements are increasingly evident in image inpainting tasks. However, many existing multi-stage methods fail to effectively inpainting the larger missing areas, their common drawback is that the result of each stage is easily misguided by the wrong content generated in the previous stage. To solve this issue, in this paper, a novel one-stage generative adversarial network based on the progressive decoding architecture and gradient guidance. Firstly, gradient priors are extracted at the encoder stage to be passed to the decoding branch, and multiscale attention fusion group is used to help the network understand the image features. Secondly, multiple parallel decoding branches fill and refine the missing regions by top-down passing the reconstructed priors. This progressively guided repair avoids the detrimental effects of inappropriate priors. The joint guidance of features and gradient priors helps the restoration results contain the correct structure and rich details. And the progressive guidance is achieved by our fusion strategy, combining reimage convolution and design channel coordinate attention to fuse and reweight the features of different branches. Finally, we use the multiscale fusion to merge the feature maps at different scales reconstructed by the last decoding branch and map them to the image space, which further improves the semantic plausibility of the restoration results. Experiments on multiple datasets show that the qualitative and quantitative results of our computationally efficient model are competitive with those of state-of-the-art methods.

Free scan real time 3D ultrasound imaging with shading artefacts removal

Ultrasound imaging (USI) is a widely adopted imaging method in clinical diagnosis owing to its low cost, convenience, and safety. However, due to the complex acoustic attenuation, two-dimensional (2D) USI lacks the capability to achieve a clear imaging result when the target is shaded by high echo tissues. This paper proposes a three-dimensional (3D) free-scan real-time ultrasound imaging (FRUSI) method. By integrating 2D ultrasound image sequences around the region of interest (ROI) with a real-time and spatially accurate probe tracking method, the proposed FRUSI system provides clear and accurate ultrasound images for medical study. The experiment results on reconstruction precision and accuracy show the potential ability of our proposed system to provide high-quality 3D ultrasound imaging. Moreover, previously shaded targets can be discerned clearly in the same scan plane in both phantom studies and in vivo studies on the human finger joint. The performance of the proposed FRUSI system has demonstrated its potential value for clinical diagnosis to provide high ultrasound imaging quality and rich details in spatial information. Due to the convenient setup, the FRUSI system might potentially be expanded to other ultrasound imaging modalities.

Context-Based Novel Histogram Bin Stretching Algorithm for Automatic Contrast Enhancement

This article presents CHBS, a novel context-based histogram bin stretching method that enhances the contrast by increasing the range of gray levels and randomness among the gray levels. It comprises image spatial contextual information and discrete cosine transform (DCT). It constitutes the global enhancement with the context-based histogram bin stretching and local details with the DCT. First, it uses the spatial similarities among surrounding pixels to generate random numbers. Unlike the other methods, the similarity map is generated based on the neighboring pixels’ mutual relationship. Intensity values are distributed among the available dynamic range to generate a global contrast-enhanced image. Second, the DCT is further applied to the previous contrast-enhanced image to adjust its local details automatically. Several experiments are conducted on the different levels of contrast degraded images. Both subjective and objective assessment outcomes validate that the projected approach is better or comparable with several state-of-the-art approaches in terms of brightness preservation, richer details, and natural appearance.

The pre-coding optimization for the digital horizontal glass-free 3D display based on holographic functional screen

Each reconstructed view perspective suffers low display quality due to the shortage of planar pixels for mapping light rays with designated intensity and directions to integrate the 3D image, which limits the 3D visual effect of the digital glass-free three-dimensional (3D) display. Here, a pre-coding optimization method of rendering the elemental image array (EIA) is further proposed and used to enhance each reconstructed view perspective with high clarity and rich details besides using the holographic functional screen (HFS). The pre-coding is realized by reassigning designated sub-pixels of the parallax images (PIs), which are used to synthesize the EIA, by linearly weighting their adjacent sub-pixels with the normalized Gaussian function before the synthetization of the EIA with the PIs. The normalized Gaussian function based reassignment of sub-pixels is consistent with the visual property of the human eye. The simulation and experimental presented 3D images with high clarity and rich details are demonstrated, verifying the feasibility and superiority of the proposed method.

A Generative Adversarial Network for infrared and visible image fusion using adaptive dense generator and Markovian discriminator

Existing infrared and visible image fusion algorithms have problems with reduced saliency for infrared targets and a lack of background texture information. To address this problem, a generative adversarial network for image fusion based on an attention mechanism is proposed in this study. This network consists of a generator, a discriminator, and an adaptive decision block. The generator employs a DenseNet-based cascade approach, which substantially lowers the information loss during the convolution process. A Markovian discriminator, unlike global discriminators, is embedded in the network. It divides images into patches, causing the network to focus on local regions and distinguishing the generated images as a whole. In addition, an adaptive decision block is included in the model to build an intensity dynamic mapping mechanism. To guide the computation for a certain content loss, a weighted map is created based on the intensity information of infrared images. Unlike the other loss functions, the weighted parameters of each component of the loss function are adaptively evaluated. To demonstrate the effectiveness of the proposed method, qualitative and quantitative experiments on TNO and RoadScene datasets are carried out. The experimental findings suggest that the proposed technique retains more significant characteristics in source images and produces fused images with rich texture details. Furthermore, the system noise resistance of each approach is examined. Under the condition of different noise, the fusion results of our approach still exhibit less distortion and a higher peak signal-to-noise ratio when compared to other state-of-the-art fusion methods.

Memory-efficient multi-scale residual dense network for single image rain removal

Outdoor photography is greatly affected by bad rainfall, which degrades the quality of captured images and hinders practical application. In response to the above problems, we propose a memory-efficient multi-scale residual dense network (MMRD-Net) for single image rain removal. Firstly, we use a combination or fusion of extended convolution blocks of different sizes to accurately extract multi-scale features of rain streaks and raindrops while retaining rich local detail. Secondly, we add attention modules to the residual block, which performs adaptive operations on rainfall and non-rainfall areas to extract more and more accurate image features and details of rain streaks and raindrops. During this period, we opened up a space for shared use by all the concatenation layers in the network, which reduces the memory consumption of the network. It allows for better aggregation of stormwater features at multiple levels and also allows for faster convergence of the network. Finally, we tested on both synthetic and real datasets to obtain subjective results and objective evaluations. The experimental results show that the performance of our algorithm is good and robust for joint rain removal tasks of different densities. Our algorithm outperforms twelve other different classical algorithms in terms of its ability to remove rain streaks. It can produce visually sharper deraining images and the enhanced visibility of the images is effective for computer vision applications (Google Vision API).

HyperTDP-Net: A Hyper-densely Connected Compression-and-Decomposition Network Based on Trident Dilated Perception for PET and MRI Image Fusion

Objective. Since medical images generated by medical devices have low spatial resolution and quality, fusion approaches on medical images can generate a fused image containing a more comprehensive range of different modal features to help physicians accurately diagnose diseases. Conventional methods based on deep learning for medical image fusion usually extract only local features without considering their global features, which often leads to the problem of unclear detail information in the final fused image. Therefore, medical image fusion is a challenging task of great relevance. Approach. This paper proposes a novel end-to-end medical image fusion model for PET and MRI images to achieve information interaction between different pathways, termed as hyper-densely connected compression-and-decomposition network based on trident dilated perception for PET and MRI image fusion (HyperTDP-Net). In particular, in the compression network, a dual residual hyper densely module is constructed to take full advantage of middle layer information. Moreover, we establish the trident dilated perception module to precisely determine the location information of features, and improve the feature representation capability of the network. In addition, we abandon the ordinary mean square error as the content loss function and propose a new content-aware loss consisting of structural similarity loss and gradient loss, so that the fused image not only contains rich texture details but also maintains sufficient structural similarity with the source images.Main results. The experimental dataset used in this paper is derived from multimodal medical images published by Harvard Medical School. Extensive experiments illustrate that our model contains more edge information and texture detail information in the fusion result than the 12 state-of-the-art fusion models and ablation study results demonstrate the effectiveness of three technical innovations.Significance. As medical images continue to be used in clinical diagnosis, our method is expected to be a tool that can effectively improve the accuracy of physician diagnosis and automatic machine detection.

DINet: Deformation Inpainting Network for Realistic Face Visually Dubbing on High Resolution Video

For few-shot learning, it is still a critical challenge to realize photo-realistic face visually dubbing on high-resolution videos. Previous works fail to generate high-fidelity dubbing results. To address the above problem, this paper proposes a Deformation Inpainting Network (DINet) for high-resolution face visually dubbing. Different from previous works relying on multiple up-sample layers to directly generate pixels from latent embeddings, DINet performs spatial deformation on feature maps of reference images to better preserve high-frequency textural details. Specifically, DINet consists of one deformation part and one inpainting part. In the first part, five reference facial images adaptively perform spatial deformation to create deformed feature maps encoding mouth shapes at each frame, in order to align with input driving audio and also the head poses of input source images. In the second part, to produce face visually dubbing, a feature decoder is responsible for adaptively incorporating mouth movements from the deformed feature maps and other attributes (i.e., head pose and upper facial expression) from the source feature maps together. Finally, DINet achieves face visually dubbing with rich textural details. We conduct qualitative and quantitative comparisons to validate our DINet on high-resolution videos. The experimental results show that our method outperforms state-of-the-art works.

Reconciling theories on why employees of small firms are more likely to become entrepreneurs

Abstract A common finding in the literature on heterogeneity in entrepreneurial entry is that workers of small firms relative to large firms are more likely to become entrepreneurs (i.e., the “small firm effect”). Researchers disagree, however, on how to interpret this relationship, and in which contexts. A traditional argument is that work experience in small firms facilitates the acquisition of knowledge relevant to entrepreneurship, which, in turn, facilitates transitions into entrepreneurship. An alternative explanation argues that, at least in technical domains, the small firm effect derives from the potential for large firms to better integrate ideas internally (i.e., intrapreneurship). We inform this debate by leveraging a unique, nationally representative survey that provides rich details about the emergence of entrepreneurial ideas and their potential implementation inside vs. outside of firms. Consistent with the former explanation of the small firm effect, for the average worker, we find strong evidence that small firms (relative to large firms) are better at providing an environment that facilitates transitions to entrepreneurship. However, we also find that technical workers are more likely to attempt to implement ideas in larger firms relative to non-tech workers. Our results thus imply that while the traditional mechanism drives the small firm effect for non-tech workers, both mechanisms contribute to the effect for the tech workers, thus helping to reconcile competing explanations in prior studies. In addition to unpacking the drivers of the small firm effect, our study has broader implications for the recent discussion on the drivers of entrepreneurial dynamism.

MGFCTFuse: A Novel Fusion Approach for Infrared and Visible Images

Traditional deep-learning-based fusion algorithms usually take the original image as input to extract features, which easily leads to a lack of rich details and background information in the fusion results. To address this issue, we propose a fusion algorithm, based on mutually guided image filtering and cross-transmission, termed MGFCTFuse. First, an image decomposition method based on mutually guided image filtering is designed, one which decomposes the original image into a base layer and a detail layer. Second, in order to preserve as much background and detail as possible during feature extraction, the base layer is concatenated with the corresponding original image to extract deeper features. Moreover, in order to enhance the texture details in the fusion results, the information in the visible and infrared detail layers is fused, and an enhancement module is constructed to enhance the texture detail contrast. Finally, in order to enhance the communication between different features, a decoding network based on cross-transmission is designed within feature reconstruction, which further improves the quality of image fusion. In order to verify the advantages of the proposed algorithm, experiments are conducted on the TNO, MSRS, and RoadScene image fusion datasets, and the results demonstrate that the algorithm outperforms nine comparative algorithms in both subjective and objective aspects.

CONNECTING DESIGN ITERATIONS TO PERFORMANCE IN ENGINEERING DESIGN

AbstractNo matter a system's size, complexity, or domain, iterations are fundamental to its design process. However, there is a duality: iterations are both signs of usefully exploring the system's design space and failure to find an appropriate solution. This ambiguity means that we have not been able to connect teams’ iterating behavior to their design's performance, potentially obscuring a way to influence the design process.As such, our exploratory study unpacks the relationship between design iterations and performance. We observed 88 teams in the 2020 Robots to the Rescue Competition in rich detail. Using logs of 7,956 iterations on a Computer-Aided Design platform, we analyzed how high- and low-performing teams revised their submissions, searching for consistent differences in their behavior. We found significant differences in the iterations’ number, scale, and cadence between these groups of teams. These findings emphasized the correlation between certain iteration patterns and the success of a design: the best teams will likely revise differently than the worst ones. It also showed the importance of a fine-grained, time- dependent view of the design process to resolve open questions in the literature.

Learning the degradation distribution for medical image superresolution via sparse swin transformer

High-resolution (HR) medical images can provide rich details, which are important for discovering subtle lesions to make diagnoses. Convolutional neural networks (CNNs) are widely used in this field, but struggle to model long-range dependencies. Although transformer-based methods have improved in this respect, this method requires large quantities of data. Unfortunately, large quantities of low-resolution (LR) and HR medical image pairs may not always be available. In addition, most medical image superresolution (SR) methods are deterministic, while the degradation in real scenarios is stochastic. To address these problems, we introduce a probabilistic degradation model that combines natural and medical images for training. This design alleviates the problem of insufficient medical image pairs and learns the degradation process of the natural scene. In addition, we propose a new medical image SR model that consists of CNNs and the Swin Transformer structure to excavate both local and global semantic features. Moreover, to reduce computational stress, the spherical locality-sensitive hashing (SLSH) module is employed in the nonlocal attention (NLA) mechanism to form the ENLA module. This design enables the proposed Sparse Swin Transformer (SSFormer) model to generate HR medical images without extensive training images. Experiments on diverse datasets (natural images and medical images) demonstrate that the proposed method is robust and effective, qualitatively and quantitatively outperforming other medical image SR methods. Code is available at https://github.com/codehxj/SSFormer.

A Comprehensive Evaluation of Three Global Surface Longwave Radiation Products

Surface longwave radiation is sensitive to climate change on Earth. This study first comprehensively evaluates the accuracies of surface longwave upward radiation (SLUR) and surface longwave downward radiation (SLDR) among the mainstream surface longwave (LW) radiation products (GLASS, CERES SYN and ERA5); then, the global annual mean values of surface LW radiation as well as its temporal variations from 2003 to 2020 are quantified. The ERA5 SLUR and SLDR show the best accuracies by direct validation, with biases/Stds/RMSEs of −1.05/18.34/18.37 W/m2 and −9.41/24.15/25.92 W/m2, respectively. The GLASS SLUR has the best accuracy under clear-sky conditions with a bias/Std/RMSE of −6.73/14.21/15.72 W/m2. The accuracy of the GLASS SLDR is comparable to CERES SYN. The merit of the GLASS LW radiation is that it can provide rich spatial details due to its high spatial resolution. The global annual mean SLUR is 399.77/398.92/398.19 W/m2, and that of the SLDR is 342.64/347.98/340.47 W/m2 for GLASS, CERES SYN and ERA5, respectively. The interannual variation trends for the three products produce substantially growing long-term trends for the global mean SLUR and SDLR over the globe and land, while there are almost no trends over the ocean. The long-term trends of the seasonal mean SLUR and SDLR in the Northern and Southern Hemispheres are asymmetrical. Our comprehensive evaluation and trend analysis of the mainstream surface LW radiation products can aid in understanding the global energy balance and climate change.

Impulse Fluid Simulation

We propose a new incompressible Navier-Stokes solver based on the impulse gauge transformation. The mathematical model of our approach draws from the impulse-velocity formulation of Navier-Stokes equations, which evolves the fluid impulse as an auxiliary variable of the system that can be projected to obtain the incompressible flow velocities at the end of each time step. We solve the impulse-form equations numerically on a Cartesian grid. At the heart of our simulation algorithm is a novel model to treat the impulse stretching and a harmonic boundary treatment to incorporate the surface tension effects accurately. We also build an impulse PIC/FLIP solver to support free-surface fluid simulation. Our impulse solver can naturally produce rich vortical flow details without artificial enhancements. We showcase this feature by using our solver to facilitate a wide range of fluid simulation tasks including smoke, liquid, and surface-tension flow. In addition, we discuss a convenient mechanism in our framework to control the scale and strength of the turbulent effects of fluid.

A Multitemporal Scale and Spatial–Temporal Transformer Network for Temporal Action Localization

Temporal action localization plays an important role in video analysis, which aims to localize and classify actions in untrimmed videos. Previous methods often predict actions on a feature space of a single temporal scale. However, the temporal features of a low-level scale lack sufficient semantics for action classification, while a high-level scale cannot provide the rich details of the action boundaries. In addition, the long-range dependencies of video frames are often ignored. To address these issues, a novel multitemporal-scale spatial–temporal transformer (MSST) network is proposed for temporal action localization, which predicts actions on a feature space of multiple temporal scales. Specifically, we first use refined feature pyramids of different scales to pass semantics from high-level scales to low-level scales. Second, to establish the long temporal scale of the entire video, we use a spatial–temporal transformer encoder to capture the long-range dependencies of video frames. Then, the refined features with long-range dependencies are fed into a classifier for coarse action prediction. Finally, to further improve the prediction accuracy, we propose a frame-level self-attention module to refine the classification and boundaries of each action instance. Most importantly, these three modules are jointly explored in a unified framework, and MSST has an anchor-free and end-to-end architecture. Extensive experiments show that the proposed method can outperform state-of-the-art approaches on the THUMOS14 dataset and achieve comparable performance on the ActivityNet1.3 dataset. Compared with A2Net (TIP20, Avg{0.3:0.7}), Sub-Action (CSVT2022, Avg{0.1:0.5}), and AFSD (CVPR21, Avg{0.3:0.7}) on the THUMOS14 dataset, the proposed method can achieve improvements of 12.6%, 17.4%, and 2.2%, respectively.

Natural Ventilation in Low-Cost Housing: An Evaluation by CFD

With the spread of sustainability concepts, due to the global energy crisis and the unrestrained consumption of natural resources, the importance of rational use and reduction of energy consumption has been intensified, albeit in short steps. This concern has taken up bioclimatic concepts in buildings, especially housing buildings for low-income people, aiming at improving the quality of housing production in terms of its habitability. The objective of this research is to theoretically and experimentally study the natural ventilation in a housing building, which fits into the low-cost pattern. The use of Computational Fluid Dynamics to predict natural ventilation was adopted as a tool to perform these analyses. Experiments were carried out by visiting the study area to collect data on air velocity, temperature, and relative humidity in all rooms of the residence in order to carry out the intended analyses. Computer simulation of natural ventilation considering the building with the doors and windows open and with fixed geometry was performed. Results proved that the proposed mathematical model was able to reproduce, with rich details and physical coherence, the internal and external air flow inside the building, indicating better internal ventilation performance in environments that have door and window openings, with recesses, revealing the importance of cross ventilation to reduce the internal temperature and consequent improvement in thermal comfort. The idea is to help civil engineers and specialists in the economically viable design of low-cost buildings from the economic, social, and thermal comfort points of view.

Machine Learning–Based Time in Patterns for Blood Glucose Fluctuation Pattern Recognition in Type 1 Diabetes Management: Development and Validation Study

Background Continuous glucose monitoring (CGM) for diabetes combines noninvasive glucose biosensors, continuous monitoring, cloud computing, and analytics to connect and simulate a hospital setting in a person’s home. CGM systems inspired analytics methods to measure glycemic variability (GV), but existing GV analytics methods disregard glucose trends and patterns; hence, they fail to capture entire temporal patterns and do not provide granular insights about glucose fluctuations. Objective This study aimed to propose a machine learning–based framework for blood glucose fluctuation pattern recognition, which enables a more comprehensive representation of GV profiles that could present detailed fluctuation information, be easily understood by clinicians, and provide insights about patient groups based on time in blood fluctuation patterns. Methods Overall, 1.5 million measurements from 126 patients in the United Kingdom with type 1 diabetes mellitus (T1DM) were collected, and prevalent blood fluctuation patterns were extracted using dynamic time warping. The patterns were further validated in 225 patients in the United States with T1DM. Hierarchical clustering was then applied on time in patterns to form 4 clusters of patients. Patient groups were compared using statistical analysis. Results In total, 6 patterns depicting distinctive glucose levels and trends were identified and validated, based on which 4 GV profiles of patients with T1DM were found. They were significantly different in terms of glycemic statuses such as diabetes duration (P=.04), glycated hemoglobin level (P<.001), and time in range (P<.001) and thus had different management needs. Conclusions The proposed method can analytically extract existing blood fluctuation patterns from CGM data. Thus, time in patterns can capture a rich view of patients’ GV profile. Its conceptual resemblance with time in range, along with rich blood fluctuation details, makes it more scalable, accessible, and informative to clinicians.