Search

Abstract Oxide-supported metal catalysts are essential components in industrial processes for catalytic conversion. However, the performance of these catalysts is often compromised in high temperature reaction environments due to sintering effects. Currently, a number of studies are underway with the objective of improving the metal support interaction (MSI) effect in order to enhance sintering resistance by surface modification of the oxide support, including the formation of inhomogeneous defects on the oxide support, the addition of a rare earth element, the use of different facets, encapsulation, and other techniques. The recent developments in in situ gas phase transmission electron microscopy (TEM) have enabled direct observation of the sintering process of NPs in real time. This capability further allows to verify the efficacy of the methods used to tailor the support surface and contributes effectively to improving sintering resistance. Here, we review a few selected studies on how in situ gas phase TEM has been used to prevent the sintering of catalyst NPs on oxide supports.

Recent advancements in blade‐coating organic photovoltaic (OPV) devices utilizing eco‐friendly nonhalogenated solvents have demonstrated high power conversion efficiencies (PCEs) when processed at high substrate temperatures. However, this method poses challenges in device reproducibility and stability. Herein, a BTP‐eC9‐γ nonfullerene acceptor (analogous to BTP‐eC9) with γ‐position‐branched inner side chains within the BTP‐eC9‐based structural motif is developed. This pin‐sized extension in the branching position enhances the solubility of BTP‐eC9‐γ in nonhalogenated toluene solvent. This improvement not only mitigates excessive aggregation in the film state but also facilitates device fabrication at lower substrate temperatures. Optimized at a substrate temperature of 40 °C, the BTP‐eC9‐γ‐based blade‐coating devices with toluene achieve remarkable PCEs of 16.43% (0.04 cm2) and 14.95% (1.0 cm2). Furthermore, these devices retain their high film uniformity at 40 °C, which contributes to superior device reproducibility. This is attributed to the minimized alteration in the evolution kinetics of fluid flow. These findings signify a promising direction for the industrial production of blade‐coating OPV devices.

Weakly-supervised semantic segmentation (WSSS) methods via transformer have been actively studied by leveraging their strong capability to capture the global context. However, since the activation function only highlights a few tokens in the self-attention mechanism of the transformer, these methods still suffer from the sparse attention map, which leads to the generation of incomplete pseudo labels. In this paper, we propose a novel class activation scheme that is able to uniformly highlight the whole object region. The key idea of the proposed method is to activate the object region by following the guide of clusters, which are formed by combining similar image features of the object. Specifically, the clustering-guided class activation map (ClusterCAM) is generated from the proposed clustering-based attention module, and highly responsive regions in this map are then adopted to activate target objects in the encoded feature space. This helps the model to explore the entire region of the target object by using the semantic proximity between patch tokens extracted from the same object. Based on this, we design an end-to-end WSSS framework that can simultaneously train classification and segmentation networks in a single-stage manner. Experimental results on benchmark datasets show that our proposed method significantly outperforms previous WSSS methods, including several multi-stage approaches. The code and model are publicly available at: https://github.com/DCVL-WSSS/ClusterCAM.

Abstract This paper proposes a modification to Tanner's work for constructing girth‐8 quasi‐cyclic low‐density parity‐check codes. The main contribution of this paper is to use an arithmetic sequence at the leftmost column for the exponent matrix so that the lifting size is not necessarily restricted to the prime numbers. Two theorems on the lifting sizes that achieve girth at least 8 using this approach is also provided. This construction exhibits better frame error rate results to the modified 5G new radio (NR) low‐density parity‐check codes for lengths around 500. Also, this construction achieves better frame error rate performance results than the recently proposed one using the Golomb rulers at around frame error rate of 10−6.

In this study, Saccharomyces cerevisiae GILA 118 was used to separate cell wall components and protoplast to confirm the alleviation of inflammation using murine macrophage cells, RAW 264.7 cells, and C57BL/6J mice compared to the commercial probiotic strain Saccharomyces boulardii. In vitro and in vivo inflammation models were treated with S. cerevisiae GILA 118 cell wall polysaccharide and protoplast to verify the alleviating inflammation. Nitric oxide was reduced, cyclo-oxygenase-2 gene expression was reduced by peroxisome proliferator-activated receptor-γ expression in RAW 264.7 cells, and S. cerevisiae GILA 118 cell wall polysaccharide treated with zymolyase (PZ) significantly increased interleukin-10 compared to the dextran sulfate sodium mouse model. S. cerevisiae GILA 118 PZ induced Treg cell proportion and decreased Th 17/Treg cells proportion in the spleen. Thus, this study suggested that yeast cell wall components are potential alleviative agents in patients with inflammatory bowel disease.

Semantic segmentation has been actively studied with the great success of deep learning. Various network architectures for this task have been introduced over last few years, and the current state-of-the-art methods have shown significant advances enough to be applied to practical applications. However, previous methods often fail to segment small objects due to the limit on the spatial resolution of the feature map encoded through the deep network architecture. In this paper, we propose a simple yet effective module for semantic segmentation. Its key components include feature high-boosting, which efficiently highlights boundaries of small objects by simple operations, i.e., sum of the original feature and the corresponding residual, in the latent space. This is fairly desirable to guide network parameters to focus on features of small objects. Experimental results on various benchmarks show that the proposed method successfully improves the performance of semantic segmentation regardless of backbone networks. In particular, the proposed method with the transformer-based decoder achieves 56.4 mIoU on the ADE20K dataset.

Low-light image enhancement is a key prerequisite for diverse applications in the field of image processing and computer vision. Various approaches for this task have been introduced over last few decades, and the current state of the art methods have shown remarkable advances based on deep neural networks. However, there are still technical issues to be resolved, e.g., dependency on subjective re-touching results and inconsistency with subjective evaluations. The goal of this work is to provide a comprehensive overview and a practical guide for experts as well as beginners. This paper covers a systematic taxonomy of existing algorithms, representative methodologies, and the performance analysis on benchmark datasets. To pave the way of the development direction for low-light image enhancement, constructive discussions and prospects are also provided.

African swine fever (ASF), a highly contagious and severe hemorrhagic viral disease in swine, is emerging as a major threat not only in Korea but also worldwide. The first confirmed case of ASF in Korea was reported in 2019. Despite the occurrence of ASF in Korea, only a few studies have genetically characterized the causative ASF virus (ASFV). In this study, we aimed to genetically characterize the ASFV responsible for the 2019 outbreak in Korea. The genome of the ASFV isolated during the first outbreak in Korea was analyzed. The Korea/YC1/2019 strain has 188,950 base pairs, with a GC content of 38.4%. The complete genome sequence was compared with other ASFV genomes annotated in the NCBI database. The Korea/YC1/2019 strain shared the highest similarity with Georgia 2007, Belgium 2018/1, and ASFV-wbBS01 strains. This study expands our knowledge of the genetic diversity of ASFV, providing valuable information for epidemiology, diagnostics, therapies, and vaccine development.

Low-light image enhancement is a key technique to overcome the quality degradation of photos taken under challenging illumination conditions. Even though the significant progress has been made for enhancing the poor visibility, the intrinsic noise amplified in low-light areas still remains as an obstacle for further improvement in visual quality. In this paper, a novel and simple method for low-light image enhancement is proposed. Specifically, the subspace, which has an ability to separately reveal illumination and noise, is constructed from a group of similar image patches, so-called volume, at each pixel position. Based on the principal energy analysis onto this volume-based subspace, the illumination component is accurately inferred from a given image while the unnecessary noise is simultaneously suppressed. This leads to clearly unveiling the underlying structure in low-light areas without loss of details. Experimental results show the efficiency and robustness of the proposed method for low-light image enhancement compared to state-of-the-art methods.

As biometric authentication systems are popularly used in various mobile devices, e.g., smart-phones and tablets, face anti-spoofing methods have been actively developed for the high-level security. However, most previous approaches still suffer from diverse types of spoofing attacks, which are hardly covered by the limited number of training datasets, and thus they often show the poor accuracy when unseen samples are given for the test. To address this problem, a novel method for face anti-spoofing is proposed based on one-class (i.e., live face only) learning with the live correlation loss. Specifically, encoder-decoder networks are firstly trained with only live faces to extract latent features, which have an ability to compactly represent various live facial properties in the embedding space and produce the spoofing cues, which are simply obtained by subtracting the original RGB image and the generated one. After that, such features are fed into the proposed feature correlation network (FCN) so that weights of FCN learn to compute “liveness” of given features under the guidance of the live correlation loss. It is noteworthy that the proposed method only requires live facial images for training the model, which are easier to obtain than fake ones, and thus the generality power for resolving the problem of face anti-spoofing can be expected to be improved. Experimental results on various benchmark datasets demonstrate the efficiency and robustness of the proposed method.

A vast amount of pictures are taken every day by using cameras mounted on various mobile devices. Even though the clarity of such acquired images has been significantly improved due to the advance of the image sensor technology, the visual quality is hardly guaranteed under varying illumination conditions. In this paper, a novel yet simple method for low-light image enhancement is proposed via the maximal diffusion value. The key idea of the proposed method is to estimate the illumination component, which is likely to appear as the bright pixel even under the low-light condition, by exploring multiple diffusion spaces. Specifically, the illumination component can be accurately separated from the scene reflectance by selecting the maximal value at each pixel position of those diffusion spaces, and thus independently adjusted for the visual quality enhancement. That is, we propose to adopt the maximal value among diffused intensities at each pixel position, so-called maximal diffusion value, as the illumination component since illumination components buried in the dark tend to be revealed with bright intensities through the iterative diffusion process. In contrast to previous approaches that still pose difficulties to balance between over-saturated and conservative restorations, the proposed method improves the image quality without any significant distortion while successfully suppressing the problem of noise amplification. Experimental results on benchmark datasets show the efficiency and robustness of the proposed method compared to previous approaches introduced in literature.

This paper addresses the problem of monocular depth estimation, which plays a key role to understand a given scene. Owing to the success of the generative model using deep neural networks, the performance of depth estimation from a single image has been significantly improved. However, most previous approaches still fail to accurately estimate the depth boundary and thus lead to the result of the blurry restoration. In this paper, a novel and simple method is proposed by exploiting the latent space of the depth-to-depth network, which contains useful encoded features for guiding the process of depth generation. This network, so-called guided network, simply consists of convolution layers and their corresponding deconvolution ones, and is also easily trained by only using single depth images. For efficiently learning the relationship between a color value and its related depth value in a given image, we propose to train the color-to-depth network via loss defined along with features from the latent space of our guided network (i.e., depth-to-depth network). One important advantage of the proposed method is to greatly enhance local details even under complicated background regions. Moreover, the proposed method works very fast (at 125 fps with GPU). Experimental results on various benchmark datasets show the efficiency and robustness of the proposed approach compared to state-of-the-art methods.

The visual quality of a captured image is often degraded by complicated lighting conditions in various real-world environments. This quality deterioration probably leads to the significant performance drop in many algorithms of computer vision, which require high-visibility inputs for precise results. In this paper, a novel method for image enhancement is proposed with the principal energy analysis. Specifically, based on the key observation that the illumination component is dominant over a small local region, the corresponding energy is efficiently separated from the scene reflectance by exploiting the subspace analysis. Owing to this clear separation, the illumination component can be easily adjusted independent of the reflectance layer for better visual aesthetics. In contrast to previous methods that still suffer from the exaggerated or conservative restoration yielding the loss of details and defects of halo artifacts, the proposed scheme has a good ability to enhance the image contrast while successfully preserving the color attribute of the original scene. Moreover, the proposed method is conceptually simple and easy to implement. Experimental results demonstrate the effectiveness of the proposed method even under diverse lighting conditions, e.g., low light, casting shadow, uneven illuminations, and so on, and the superiority of the proposed method over previous approaches introduced in the literature.

Background subtraction is a key prerequisite for a wide range of image processing applications due to its pervasiveness in various contexts. In particular, video surveillance highly requires the reliable background subtraction for further operations, such as object tracking and recognition, and thus, enormous efforts for this task have been devoted in recent decades. However, the path of technological evolution for background subtraction has now faced with an important issue that has started to be resolved: sensitivity to dynamic changes of scene contexts (e.g., illumination variations and moving backgrounds). Such dynamic changes are hardly tolerated by most of traditional background models, since they yield the drastically different statistics of pixel values even onto the relevant position between consecutive frames. To resolve this problem, many researchers in this field have developed robust and efficient methods. The goal of this paper is to provide a comprehensive review with a special attention to schemes related to handling varying illuminations frequently occurring in the outdoor surveillance scenario. This paper covers a systematic taxonomy, methodologies, and performance evaluations on benchmark databases, and also provides constructive discussions for the smart video surveillance under unconstrained outdoor environments.

Visual object tracking is an essential technique for constructing intelligent livestock management systems. Behavior patterns estimated from the trajectories of animals provide substantial useful information related to estrus cycle, disease prognosis and so on. However, similar colors and shapes between animals often lead to the failure of tracking multiple objects, and the background clutter of the breeding space further makes the problem intractable. In this paper, we propose a novel method for tracking animals using a single thermal sensor. The key idea of the proposed method is to represent the foreground (i.e., animals) easily obtained by a simple thresholding in a thermal frame as a topographic surface, which is very helpful for finding the boundary of each object even in cases with overlapping. Based on the segmentation results derived from morphological operations on the topographic surface, the center positions of all the animals are consistently updated with an efficient refinement scheme that is robust to the abrupt motions of animals. Experimental results using various thermal video sequences demonstrate the efficiency and robustness of our method for tracking animals in a breeding space compared to previous approaches proposed in the literature.

This paper presents a method for recognizing human actions from a single query action video. We propose an action recognition scheme based on the ordinal measure of accumulated motion, which is robust to variations of appearances. To this end, we first define the accumulated motion image (AMI) using image differences. Then the AMI of the query action video is resized to a N×N subimage by intensity averaging and a rank matrix is generated by ordering the sample values in the sub-image. By computing the distances from the rank matrix of the query action video to the rank matrices of all local windows in the target video, local windows close to the query action are detected as candidates. To find the best match among the candidates, their energy histograms, which are obtained by projecting AMI values in horizontal and vertical directions, respectively, are compared with those of the query action video. The proposed method does not require any preprocessing task such as learning and segmentation. To justify the efficiency and robustness of our approach, the experiments are conducted on various datasets.