Search

Blur artifacts can seriously degrade the visual quality of images, and numerous deblurring methods have been proposed for specific scenarios. However, in most real-world images, blur is caused by different factors, e.g., motion, and defocus. In this paper, we address how other deblurring methods perform in the case of multiple types of blur. For in-depth performance evaluation, we construct a new large-scale multi-cause image deblurring dataset (MC-Blur), including real-world and synthesized blurry images with different blur factors. The images in the proposed MC-Blur dataset are collected using other techniques: averaging sharp images captured by a 1000-fps high-speed camera, convolving Ultra-High-Definition (UHD) sharp images with large-size kernels, adding defocus to images, and real-world blurry images captured by various camera models. Based on the MC-Blur dataset, we conduct extensive benchmarking studies to compare SOTA methods in different scenarios, analyze their efficiency, and investigate the buildataset’s capacity. These benchmarking results provide a comprehensive overview of the advantages and limitations of current deblurring methods, revealing our dataset’s advances. The dataset is available to the public at https://github.com/HDCVLab/MC-Blur-Dataset.

By hiding the front-facing camera below the display panel, Under-Display Camera (UDC) provides users with a full-screen experience. However, due to the characteristics of the display, images taken by UDC suffer from significant quality degradation. Methods have been proposed to tackle UDC image restoration and advances have been achieved. There are still no specialized methods and datasets for restoring UDC face images, which may be the most common problem in the UDC scene. To this end, considering color filtering, brightness attenuation, and diffraction in the imaging process of UDC, we propose a two-stage network UDC Degradation Model Network named UDC-DMNet to synthesize UDC images by modeling the processes of UDC imaging. Then we use UDC-DMNet and high-quality face images from FFHQ and CelebA-Test to create UDC face training datasets FFHQ-P/T and testing datasets CelebA-Test-P/T for UDC face restoration. We propose a novel dictionary-guided transformer network named DGFormer. Introducing the facial component dictionary and the characteristics of the UDC image in the restoration makes DGFormer capable of addressing blind face restoration in UDC scenarios. Experiments show that our DGFormer and UDC-DMNet achieve state-of-the-art performance.

This paper presents a comprehensive investigation into the low cycle fatigue (LCF) and creep-fatigue interaction (CFI) behavior of 2.25CrMoV steel at an elevated temperature of 455 °C. The LCF tests are conducted using a triangular waveform with varying strain amplitudes ranging from 0.4 % to 0.7 %. Furthermore, the CFI tests are conducted using a trapezoidal waveform with different hold times and hold directions. To gain further insights, the microstructure and fracture behavior of the steel are characterized using optical, scanning, and transmission electron microscopy techniques. The results reveal that the softening behavior and fatigue life degradation of the steel are dependent on the applied strain amplitude, hold time, and hold direction. Higher strain amplitudes in the LCF tests lead to an increased number of crack initiation sources. During the CFI tests, fatigue fracture is identified as the primary failure mechanism under tensile hold conditions, and an increase in hold time promotes crack propagation. The interaction between fatigue and creep is found to be more significant in compressive hold and tensile compressive combination hold conditions. Microstructure analysis demonstrates that the lath structure recovers during cyclic loading, resulting in cyclic softening. Finally, a modified energy-based model is proposed to distinguish the different roles of tensile energy and compressive energy in the fatigue process. The proposed model accurately predicts the fatigue life of the 2.25CrMoV steel, as demonstrated by the good agreement between the experimental and predicted results.

Adverse weather conditions, such as rain, raindrop, snow and haze, consistently degrade images in an unpredictable manner, thereby rendering existing task-specific and task-aligned methods inadequate in addressing this formidable problem. To this end, we investigate the application of Transformer in image restoration and introduce an efficient frequency-oriented method called AIRFormer, which is designed to restore weather-degraded images comprehensively and holistically. Specifically, we identify that the initial self-attention mechanism exhibits distinctive properties akin to a low-pass filter. Therefore, we construct a frequency-guided Transformer encoder by incorporating wavelet-based prior information to guide the extraction of image features. Additionally, considering the non-specific frequency characteristics of self-attention in the later stages, we develop a frequency-refined Transformer decoder that incorporates learnable task-specific queries across spatial dimensions, channel dimensions, and wavelet domains. To facilitate the training of our proposed method, we curate a comprehensive benchmark dataset named AIR40K that, encompasses a wide range of challenging scenarios. Extensive experimental evaluations demonstrate the superiority of our AIRFormer over both task-aligned and all-in-one methods across 15 publicly available datasets. Notably, AIRFormer achieves the best trade-off between the inference time and quality of reconstructed image, comparing with existing methods such as TransWeather and Restormer. The source code, dataset and pre-trained models will be available at https://github.com/chdwyb/AIRFormer.

Vision transformers have shown great success on numerous computer vision tasks. However, their central component, softmax attention, prohibits vision transformers from scaling up to high-resolution images, due to both the computational complexity and memory footprint being quadratic. Linear attention was introduced in natural language processing (NLP) which reorders the self-attention mechanism to mitigate a similar issue, but directly applying existing linear attention to vision may not lead to satisfactory results. We investigate this problem and point out that existing linear attention methods ignore an inductive bias in vision tasks, i.e., 2D locality. In this article, we propose Vicinity Attention, which is a type of linear attention that integrates 2D locality. Specifically, for each image patch, we adjust its attention weight based on its 2D Manhattan distance from its neighbouring patches. In this case, we achieve 2D locality in a linear complexity where the neighbouring image patches receive stronger attention than far away patches. In addition, we propose a novel Vicinity Attention Block that is comprised of Feature Reduction Attention (FRA) and Feature Preserving Connection (FPC) in order to address the computational bottleneck of linear attention approaches, including our Vicinity Attention, whose complexity grows quadratically with respect to the feature dimension. The Vicinity Attention Block computes attention in a compressed feature space with an extra skip connection to retrieve the original feature distribution. We experimentally validate that the block further reduces computation without degenerating the accuracy. Finally, to validate the proposed methods, we build a linear vision transformer backbone named Vicinity Vision Transformer (VVT). Targeting general vision tasks, we build VVT in a pyramid structure with progressively reduced sequence length. We perform extensive experiments on CIFAR-100, ImageNet-1 k, and ADE20 K datasets to validate the effectiveness of our method. Our method has a slower growth rate in terms of computational overhead than previous transformer-based and convolution-based networks when the input resolution increases. In particular, our approach achieves state-of-the-art image classification accuracy with 50% fewer parameters than previous approaches.

The interfacial interaction of 2D materials with the substrate leads to striking surface faceting affecting its electronic properties. Here, we quantitatively study the orientation-dependent facet topographies observed on the catalyst under graphene using electron backscatter diffraction and atomic force microscopy. The original flat catalyst surface transforms into two facets: a low-energy low-index surface, e.g. (111), and a vicinal (high-index) surface. The critical role of graphene strain, besides anisotropic interfacial energy, in forming the observed topographies is revealed by molecular simulations. These insights are applicable to other 2D/3D heterostructures.

Video deraining is an important task in computer vision as the unwanted rain hampers the visibility of videos and deteriorates the robustness of most outdoor vision systems. Despite the significant success which has been achieved for video deraining recently, two major challenges remain: 1) how to exploit the vast information among successive frames to extract powerful spatio-temporal features across both the spatial and temporal domains, and 2) how to restore high-quality derained videos with a high-speed approach. In this paper, we present a new end-to-end video deraining framework, dubbed Enhanced Spatio-Temporal Interaction Network (ESTINet), which considerably boosts current state-of-the-art video deraining quality and speed. The ESTINet takes the advantage of deep residual networks and convolutional long short-term memory, which can capture the spatial features and temporal correlations among successive frames at the cost of very little computational resource. Extensive experiments on three public datasets show that the proposed ESTINet can achieve faster speed than the competitors, while maintaining superior performance over the state-of-the-art methods. https://github.com/HDCVLab/Enhanced-Spatio-Temporal-Interaction-Learning-for-Video-Deraining.

Recent deep face hallucination methods show stunning performance in super-resolving severely degraded facial images, even surpassing human ability. However, these algorithms are mainly evaluated on non-public synthetic datasets. It is thus unclear how these algorithms perform on public face hallucination datasets. Meanwhile, most of the existing datasets do not well consider the distribution of races, which makes face hallucination methods trained on these datasets biased toward some specific races. To address the above two problems, in this paper, we build a public Ethnically Diverse Face dataset, EDFace-Celeb-1 M, and design a benchmark task for face hallucination. Our dataset includes 1.7 million photos that cover different countries, with relatively balanced race composition. To the best of our knowledge, it is the largest-scale and publicly available face hallucination dataset in the wild. Associated with this dataset, this paper also contributes various evaluation protocols and provides comprehensive analysis to benchmark the existing state-of-the-art methods. The benchmark evaluations demonstrate the performance and limitations of state-of-the-art algorithms. https://github.com/HDCVLab/EDFace-Celeb-1M.

Metal additive manufacturing (AM) will impact many industries, including those producting of advanced structural, propulsion, electronic, and thermal systems. Many established metal AM processes involve layer-by-layer deposition and selective binding or melting of metal powders. However, these processes do not easily permit multi-material printing or printing directly onto non-planar surfaces. By contrast, AM techniques that deposit metal directly are typically low resolution (e.g., directed energy deposition), or cannot achieve bulk metal properties (e.g., ink-based methods). Here, we present a high resolution direct metal printing method that potentially addresses these limitations. Individual metal microparticles are electrohydrodynamically ejected on-demand from a nozzle-confined meniscus and laser-melted in-flight before landing and solidifying on a substrate. We demonstrate printing of solder and platinum particles ranging in size from 30 to 150 µm and explore the process parameter space as limited by the ejection conditions and the kinetics of melting, impact, and solidification. Our experiments demonstrate process compatibility with a wide range of powder feedstock materials, and tunability of the solidified morphology via the laser processing parameters.

In patients with pre-existing heart failure, subclinical hypothyroidism or low T3 syndrome are associated with poor outcomes, and several studies have reported improved prognosis with T3 treatment. In animal models of heart failure, we have shown that low dose T3 treatment improves contractile function by attenuating adverse remodeling of the T-tubule (TT) network, and here we studied T3 effects on TT organization and function in thyroid deficiency.Methods:Female Sprague Dawley rats (~220g) ingested 0.025% PTU in drinking water to induce thyroid deficiency. Control rats had access to untreated water. After 8 weeks, a subgroup of PTU treated rats received T3 (10ug/kg/d) in PTU-containing water for an additional 2 weeks. At week 10, echocardiography and LV pressures were recorded. Ventricular myocytes were isolated by collagenase digestion, and contractile and calcium transients were recorded in single myocytes by IonOptix imaging. Live cardiomyocytes were labeled with wheat germ agglutinin (WGA)-488 or di-8-ANNEPS, and images were captured by scanning confocal microscopy and analyzed for TT periodicity. Ventricular RNA was analyzed by RT-qPCR.Results:Echocardiography and hemodynamic measurements showed improved (p<0.0001) systolic and diastolic functions with T3 treatment vs PTU (+dP/dt = 7378?1441 vs 3253?343; -dP/dt = 6900?1280 vs 2901?348). Isolated cardiomyocyte contractile measurements were impaired in PTU and significantly improved with T3 treatment (Vmax contraction = 2.54?0.52 vs 1.67?0.38; time to peak contraction (s) = 0.20?0.03 vs 0.32?0.05; Vmax relengthening = 1.55?0.59 vs 1.00?0.34). Ca2+transients showed significant increases with T3 treatment vs PTU (Vmax Ca2+release = 16.28?2.21 vs 9.13?3.06; Vmax Ca2+reuptake = 1.07?0.24 vs 0.67?0.23). TT periodicity was significantly reduced in PTU myocytes and increased (p< 0.01) with T3 treatment (TTpower = 85?3 control, 73?3 PTU and 82?4 PTU+T3). Expression of bridging integrator 1 (Bin1), integral to TT formation, was decreased in thyroid deficiency and normalized by T3.Conclusions:T3 is a critical regulator of TT structures and further supports the role of thyroid hormones in maintaining cardiac function in chronic diseases with thyroid deficiency including heart failure.