Search

Abstract Uneven lighting conditions often occur during real-life photography, such as images taken at night that may have both low-light dark areas and high-light overexposed areas. Traditional algorithms for enhancing low-light areas also increase the brightness of overexposed areas, affecting the overall visual effect of the image. Therefore, it is important to achieve differentiated enhancement of low-light and high-light areas. In this paper, we propose a network called correcting uneven illumination network (CUI-Net) with sparse attention transformer and convolutional neural network (CNN) to better extract low-light features by constraining high-light features. Specifically, CUI-Net consists of two main modules: a low-light enhancement module and an auxiliary module. The enhancement module is a hybrid network that combines the advantages of CNN and Transformer network, which can alleviate uneven lighting problems and enhance local details better. The auxiliary module is used to converge the enhancement results of multiple enhancement modules during the training phase, so that only one enhancement module is needed during the testing phase to speed up inference. Furthermore, zero-shot learning is used in this paper to adapt to complex uneven lighting environments without requiring paired or unpaired training data. Finally, to validate the effectiveness of the algorithm, we tested it on multiple datasets of different types, and the algorithm showed stable performance, demonstrating its good robustness. Additionally, by applying this algorithm to practical visual tasks such as object detection, face detection, and semantic segmentation, and comparing it with other state-of-the-art low-light image enhancement algorithms, we have demonstrated its practicality and advantages.

The expression changes of high-mobility group box 1 (HMGB1) in the mouse cochlea have recently been implicated in noise-induced hearing loss, suggesting that HMGB1 participates in regulating cochlear function. However, the precise role of HMGB1 in the auditory system remains largely unclear. This study aimed to investigate its function in the developing mouse cochlea by examining the expression pattern of HMGB1 in the mouse cochlea from embryonic day (E) 18.5 to postnatal day (P) 28 using double immunofluorescence on frozen sections. Our findings revealed that HMGB1 was extensively expressed in the cell nucleus across various regions of the mouse cochlea, including the organ of Corti. Furthermore, its expression underwent developmental regulation during mouse cochlear development. Specifically, HMGB1 was found to be localized in the tympanic border cells at each developmental stage, coinciding with the gradual anatomical in this region during development. In addition, HMGB1 was expressed in the greater epithelial ridge (GER) and supporting cells of the organ of Corti, as validated by the supporting cell marker Sox2 at P1 and P8. However, at P14, the expression of HMGB1 disappeared from the GER, coinciding with the degeneration of the GER into the inner sulcus cells. Moreover, we observed that HMGB1 co-localized with Ki-67-positive proliferating cells in several cochlear regions during late embryonic and early postnatal stages, including the GER, the tympanic border cells, cochlear lateral wall, and cochlear nerves. Furthermore, by dual-staining Ki-67 with neuronal marker TUJ1 and glial marker Sox10, we determined the expression of Ki-67 in the neonatal glial cells. Our spatial-temporal analysis demonstrated that HMGB1 exhibited distinct expression patterns during mouse cochlear development. The co-localization of HMGB1 with Ki-67-positive proliferating cells suggested that HMGB1 may play a role in cochlear development.

Field-programmable gate array (FPGA)-based real-time simulation plays a crucial role in testing power–electronic dominated systems with the formation of controller hardware-in-the-loop (CHIL) or power hardware-in-the-loop (PHIL). This work describes an efficient implementation of computation time and resource usage in the FPGA-based study of a modular multilevel converter (MMC) with detailed electromagnetic transients. The proposed modeling technique can be used in continuous control mode (CCM) and discontinuous control mode (DCM) for high-switching frequency semiconductor technologies. An FPGA-based designed solver structure is also presented to take advantage of the parallel features of FPGAs to achieve an ultra-fast calculation speed. In addition, two different switch modeling techniques are discussed with a five-level MMC case study. Experimental results on the NI PXIe platform show the feasibility of the proposed implementation, and a time step of 100 nanoseconds is achieved.

Mitogen-activated protein kinases (MAPKs and MPKs) are important in the process of resisting plant stress. In this study, 21, 12, 18, 16, and 10 MPKs were identified from Musa acuminata, Musa balbisiana, Musa itinerans, Musa schizocarpa, and Musa textilis, respectively. These MPKs were divided into Group A, B, C, and D. Phylogenetic analysis revealed that this difference in number was due to the gene shrinkage of the Group B subfamily of Musa balbisiana and Musa textilis. KEGG annotations revealed that K14512, which is involved in plant hormone signal transduction and the plant–pathogen interaction, was the most conserved pathway of the MPKs. The results of promoter cis-acting element prediction and focTR4 (Fusarium oxysporum f. sp. cubense tropical race 4) transcriptome expression analysis preliminarily confirmed that MPKs were relevant to plant hormone and biotic stress, respectively. The expression of MPKs in Group A was significantly upregulated at 4 °C, and dramatically, the MPKs in the root were affected by low temperature. miR172, miR319, miR395, miR398, and miR399 may be the miRNAs that regulate MPKs during low-temperature stress, with miR172 being the most critical. miRNA prediction and qRT-PCR results indicated that miR172 may negatively regulate MPKs. Therefore, we deduced that MPKs might coordinate with miR172 to participate in the process of the resistance to low-temperature stress in the roots of the banana. This study will provide a theoretical basis for further analysis of the mechanism of MPKs under low-temperature stress of bananas, and this study could be applied to molecular breeding of bananas in the future.

Polymetallic nodules, found abundantly in deep-ocean deposits, possess significant economic value and represent a valuable resource due to their high metal enrichment, crucial for the high-tech industry. However, accurately evaluating these valuable mineral resources presents challenges for traditional image segmentation methods due to issues like color distortion, uneven illumination, and the diverse distribution of nodules in seabed images. Moreover, the scarcity of annotated images further compounds these challenges, impeding resource assessment efforts. To overcome these limitations, we propose a novel two-stage diffusion-based model for nodule image segmentation, along with a linear regression model for predicting nodule abundance based on the coverage obtained through nodule segmentation. In the first stage, we leverage a diffusion model trained on predominantly unlabeled mineral images to extract multiscale semantic features. Subsequently, we introduce an efficient segmentation network designed specifically for nodule segmentation. Experimental evaluations conducted on a comprehensive seabed nodule dataset demonstrate the exceptional performance of our approach compared to other deep learning methods, particularly in addressing challenging conditions like uneven illumination and dense nodule distributions. Our proposed model not only extends the application of diffusion models but also exhibits superior performance in seabed nodule segmentation. Additionally, we establish a linear regression model that accurately predicts nodule abundance by utilizing the coverage calculated through seabed nodule image segmentation. The results highlight the model’s capacity to accurately assess nodule coverage and abundance, even in regions beyond the sampled sites, thereby providing valuable insights for seabed resource evaluation.

The effects of groundwater table fluctuation (GTF) on the remediation of a petrochemically polluted riverside using soil vapor extraction (SVE) were investigated. The migration and transformation of benzene, toluene, ethylbenzene, and o-xylene (BTEX) in cases of natural attenuation, SVE without GTF, and SVE with GTF were simulated using the TMVOC model. The results showed that the optimized extraction well pressure and influencing radius of the target site were 0.90 atm and 8 m, respectively. The removal rates of BTEX in cases of natural attenuation, SVE without GTF, and SVE with GTF were 11.49%, 85.16%, and 97.33%, respectively. The removal rate of BTEX was maximized in the case of SVE with a GTF amplitude of 0.5 m to 1 m. The removal rates of benzene (99.99%), toluene (99.74%), ethylbenzene (96.37%), and o-xylene (94.72%) were maximized in the case of SVE with GTF. For the cases of SVE without GTF and SVE with GTF, mass losses of BTEX in gaseous (0.05 kg, 0.05 kg, respectively) and aqueous phases (5.46 kg, 5.87 kg, respectively) were consistent. However, the mass loss of BTEX in the non-aqueous phase liquid (NAPL) phase in the case of SVE with GTF (155.13 kg) exceeded that in the case of SVE without GTF (135.41 kg). This is because GTF positively affected both the solubility and volatility of BTEX in the NAPL phase. With the groundwater table decreasing, flows of gas and gaseous pollutants increased by 25% along the vertical section. At this stage, the removal rates of volatile organic compounds can be further improved by increasing the flow of the extraction well.

Abstract Background Fibrinogen (FIB) has been found to be a promising marker in diagnosing periprosthetic joint infection (PJI), however, the value of FIB in predicting reinfection of PJI is unknown. The purpose of this study was to evaluate the value of FIB in predicting reinfection after debridement, antibiotics, and implant retention (DAIR) for PJI. Methods We retrospectively analyzed the clinical data of patients who were diagnosed with PJI and underwent DAIR from 2013 to 2019. The levels of the FIB, erythrocyte sedimentation rate (ESR), and C-reactive protein (CRP) were measured before DAIR. After DAIR, patients were followed and reinfections were identified. For both acute and chronic PJI, the predictive value of FIB was evaluated by calculating the sensitivity, specificity, and area under the curve (AUC) of the receiver operating characteristic curve (ROC), and was compared with traditional inflammatory markers including ESR and CRP. Results The expression of FIB differed between patients reinfected and those not reinfected in both acute and chronic PJI (p

Considering that the motion state and the stress of floating hose are complex under the combined action of winds, waves and currents, the dynamic simulation analysis of offshore floating hose oil transportation system was thereby carried out through the method of joint simulation based on AQWA and OrcaFlex.The numerical model was established to analyze the frequency domain of oil tanker by AQWA, and to perform the time domain analysis on the coupling motion between floating hose, oil tanker, buoy and mooring line by OrcaFlex.The results show that the influence of wind, waves and currents on the floating hose tension increases nonlinearly.The maximum tension points of the floating hose oil transportation system are at both ends of the hose.The tension acting on the hose can be obviously reduced and the drift distance can be shortened by setting securing anchors along the line.Through the analysis on the offshore floating hose oil transportation system with the transportation distance of 1 100 m, based on this method and model, it can be concluded that the optimal hose length is 1 300 m, the hose drift distance is about 300 m when no securing anchor is set along the line, and the safety requirements can be met under 4-level sea state.The research results can provide certain reference for the dynamic simulation analysis of such offshore floating hose oil transportation system.

The task of relation extraction is classifying the relations between two entities in a sentence. Distant supervision relation extraction can automatically align entities in texts based on Knowledge Base without labeled training data. For low-resource language relation extraction, such as Tibetan, the main problem is the lack of labeled training data. In this paper, we propose an improved distant supervised relation extraction model based on Piecewise Convolutional Neural Network (PCNN) to expand the Tibetan corpus. We add self-attention mechanism and soft-label method to decrease wrong labels, and use Embeddings from Language Models (ELMo) to solve the semantic ambiguity problem. Meanwhile, according to the Tibetan characteristics, we combine the word vector and part of speech vector to extract deeply feature of words. Finally, the experimental results show that P@avg value increases by 14.4% than baseline.

Violent video constitutes a threat to public security, and effective detection algorithms are in urgent need. In order to improve the detection accuracy of 3D convolutional neural networks (3D ConvNet), a novel violent video detection scheme based on the modified 3D ConvNet is proposed. In this paper, the preprocessing method of data is improved, and a new sampling method by using the key frame as dividing nodes is designed. Then, a random sampling method is adapted to produce the input frame sequence. With experimental evaluations on the crowd violence dataset, the results demonstrate the effectiveness of the proposed new sampling method. For three public violent detection datasets: hockey fight, movies, and crowd violence, individualized strategies are implemented to suit the varied clip length. For the short clips, the 3D ConvNet is constructed by using the uniform sampling method. For the longer clips, the new frame sampling strategy is adopted. The proposed scheme obtains competitive results: 99.62% on hockey fight, 99.97% on movies, and 94.3% on crowd violence. The experimental results show that our method is simple and effective.

In recent years, the large-scale open Chinese and English question answering (QA) corpora have provided important support for the application of deep learning in the Chinese and English QA systems. However, for low-resource languages, such as Tibetan, it is difficult to construct satisfactory QA systems, owing to the lack of large-scale Tibetan QA corpora. To solve this problem, this paper proposes a QA corpus generation model, called QuGAN. This model combines Quasi-Recurrent Neural Networks and Reinforcement Learning. The Quasi-Recurrent Neural Networks model is used as a generator for Generative Adversarial Network, which speeds up the generation of text. At the same time, the reward strategy and Monte Carlo search strategy are optimized to effectively update the generator network. Finally, we use the Bidirectional Encoder Representations from Transformers model to correct the generated questions at the grammatical level. The experimental results show that our model can generate a certain amount of effective Tibetan QA corpus, and the BLEU-2 value increases by 13.07% than baseline. Moreover, the speed of the model has been greatly improved.

Abstract Chitin is one of the key components of fungal cell wall, and chitin deacetylases (CDAs) have been found in fungi; however, their functions remain unknown. Aspergillus fumigatus is known to cause fatal invasive aspergillosis (IA) among immunocompromised patients with a high mortality rate. Although the A. fumigatus cell wall has long been taken as a unique target for drug development, its dynamic remodeling is complicated and not well understood. Seven putative CDAs are annotated in the A. fumigatus genome. In this study, we analyzed the function of the putative CDAs, Cod4 and Cod7, in A. fumigatus. Biochemical analysis of recombinant proteins showed that Cod4 preferentially deacetylated (GlcNAc)4 and was less active on chitooligosaccharides with DP > 5, whereas Cod7 was unable to catalyze deacetylation. Simulation of three‐dimensional structure revealed that both Cod4 and Cod7 shared a similar folding pattern with HyPgdA from Helicobacter pylori and, similar to HyPgdA, a substitution of Thr8 by Ala8 in Cod7 abolished its CDA activity. Deletion of the cod4, cod7, or both in A. fumigatus led to polarity abnormality and increased conidiation. Furthermore, the expression level of the genes related to polarity was upregulated in the mutants. Our results demonstrated that Cod4 and Cod7 were involved in polarity, though Cod4 was inactive.

With the rapid increase of vehicular Internet of things applications, it is urgent to design a mobile edge computing architecture, which is possible to distribute and process a large amount of contents with vehicles on the road. From a communication perspective, the current cellular technology faces challenges due to the limited bandwidth in a dense vehicle environment. In this paper, we propose a multi-access edge computing framework and the corresponding communication protocol which integrates licensed Sub-6 GHz band, IEEE 802.11p, and millimeter wave (mmWave) communications for the content distribution and processing in vehicular networks. The proposed protocol uses a cluster-based approach, where a fuzzy logic-based algorithm is employed to select efficient gateway nodes which bridge the licensed Sub-6 GHz communication and the mmWave communication in order to maximize the overall network throughput. IEEE 802.11p vehicle-to-vehicle communication is used to share information among vehicles in order to achieve efficient clustering. We conduct extensive simulations to evaluate the performance of the proposed protocol under various network conditions. Simulation results show that the proposed protocol can achieve significant improvements in various scenarios compared with the existing approaches.

Groundwater is an irreplaceable resource for irrigation and drinking in the North China Plain, and the quality of groundwater is of great importance to human health and social development. In this study, using the information from 59 groups of groundwater samples, groundwater quality conditions for irrigation and drinking purposes in an agricultural region of the North China Plain were analyzed. The groundwater belongs to a Quaternary loose rock pore water aquifer. The depths of shallow groundwater wells are 20–150 m below the surface, while the depths of deep groundwater wells are 150–650 m. The sodium adsorption ratio (SAR), sodium percentage (%Na), residual sodium carbonate (RSC), magnesium hazard (MH), permotic index (PI) and electrical conductivity (EC) were selected as indexes to evaluate the shallow groundwater suitability for irrigation. What’s more, the deep groundwater suitability for drinking was assessed and the human health risk of excessive chemicals in groundwater was studied. Results revealed that SAR, Na% and RSC indexes indicated the applicability of shallow groundwater for agricultural irrigation in the study area. We found 57.1% of the shallow groundwater samples were located in high salinity with a low sodium hazard zone. The concentrations of fluorine (F−) in 79.0% of the deep groundwater samples and iodine (I−) in 21.1% of the deep groundwater samples exceeded the permissible limits, respectively. The total hazard quotient (HQ) values of fluorine in over half of the deep groundwater samples exceeded the safety limits, and the health risk degree was ranked from high to low as children, adult females and adult males. In addition to natural factors, the soil layer compression caused by groundwater over-exploitation increased the fluorine concentration in groundwater. Effective measures are needed to reduce the fluorine content of the groundwater of the study area.

This study aims at developing a non-(semi-)parametric method to extract the hidden network structure from the {0,1}-valued distribution flow data with missing observations on the links between nodes. Such an input data type widely exists in the studies of information propagation process, such as the rumor spreading through social media. In that case, a social network does exist as the media of the spreading process, but its link structure is completely unobservable; therefore, it is important to make inference of the structure (links) of the hidden network. Unlike the previous studies on this topic which only consider abstract networks, we believe that apart from the link structure, different social-economic features and different geographic locations of nodes can also play critical roles in shaping the spreading process, which has to be taken into account. To uncover the hidden link structure and its dependence on the external social-economic features of the node set, a multidimensional spatial social network model is constructed in this study with the spatial dimension large enough to account for all influential social-economic factors. Based on the spatial network, we propose a nonparametric mean-field equation to govern the rumor spreading process and apply the likelihood estimator to make inference of the unknown link structure from the observed rumor distribution flows. Our method turns out easily extendible to cover the class of block networks that are useful in most real applications. The method is tested through simulated data and demonstrated on a data set of rumor spreading on Twitter.

With the development of image editing software techniques, the content integrity and authenticity of original digital images become more and more important in digital content security. A novel image tampering detection and recovery algorithm based on digital watermarking technology and a chaotic system is proposed, and it can effectively locate the tampering region and achieve the approximate recovery of the original image by using the hidden information. The pseudo-random cyclic chain is realized by the chaotic system to construct the mapping relationship between the image subblocks. It can effectively guarantee the randomness of the positional relationship between the hidden information and the original image block for the better ergodicity of the pseudo-random chain. The recovery value optimization algorithm can represent image information better. In addition to the traditional Level-1 recovery, a weight adaptive algorithm is designed to distinguish the original block from the primary recovery block, allowing 3 × 3 neighbor block recovery to achieve better results. The experimental results show that the hierarchical tamper detection algorithm makes tamper detection have higher precision. When facing collage attacks and large general tampering, it will have higher recovery image quality and better resistance performance.