Optimised CNN Architectures for Handwritten Arabic Character Recognition

Handwritten character recognition is considered challenging compared with machine-printed characters due to the different human writing styles.Arabic is morphologically rich,and its characters have a high similarity.The Arabic language includes 28 characters.Each character has up to four shapes according to its location in the word(at the beginning,middle,end,and isolated).This paper proposed 12 CNN architectures for recognizing handwritten Arabic characters.The proposed architectures were derived from the popular CNN architectures,such as VGG,ResNet,and Inception,to make them applicable to recognizing character-size images.The experimental results on three well-known datasets showed that the proposed architectures significantly enhanced the recognition rate compared to the baseline models.The experiments showed that data augmentation improved the models’accuracies on all tested datasets.The proposed model outperformed most of the existing approaches.The best achieved results were 93.05%,98.30%,and 96.88%on the HIJJA,AHCD,and AIA9K datasets.

Detection of ARV-Resistant Mutants in HIV-1-Infected Individuals in a Context of Systematic Switching to an Association Based on Dolutegravir in Abidjan, Côte d’Ivoire

The emergence of antiretroviral resistance mutations represents a major threat to the achievement of national and global goals for the elimination of HIV-1 infection. The global strategy in 2019 in Cte d'Ivoire is a new national policy for the management of people living with HIV with the administration of dolutegravir (DTG)-based fixed-dose combination. The aim of our study was to evaluate HIV-1 resistance to antiretrovirals (ARVs) in infected adult subjects in Cte d’Ivoire in the context of a systematic switch to a DTG-based combination. Between February 2022 and October 2023, a cross-sectional survey with random sampling was conducted in 06 services caring for people living with HIV. A total of 139 participants were included in the study. Adults with a viral load ≥ 1000 copies/mL were tested for HIV-1 ARV resistance mutations. Molecular analyses were performed using protocol of ANRS-MIE (National Agency for Research on AIDS and emerging infectious diseases). The interpretation is performed by HIVGRAD (https://www.hiv-grade.de/cms/grade/). The frequencies of HIV-1 resistance to non-nucleotide reverse transcriptase inhibitors (NNRTIs), nucleotide reverse transcriptase inhibitors (NRTIs), integrase inhibitors (IINTs) and protease inhibitors (PIs) were 82%, 73%, 19% and 11% respectively. The main mutations observed in the different classes were K103N (45%), M184V (64%), E157Q (19%) and L10V/M46I/A71V/I54V (6%) respectively. This study reveals the emergence of resistance to DTG-based fixed-dose combinations, favored by high rates of resistance to NRTIs and NNRTIs. This finding underlines the need for enhanced viral load monitoring and HIV-1 genotyping tests to guide the choice of NRTIs for combination therapy. In addition, monitoring for mutations to second-generation NRTIs is essential, given the scale-up of DTG-based regimens currently underway in Cte d’Ivoire.

Effect of Co on Solidification Characteristics and Microstructural Transformation of Non-equilibrium Solidified Cu-Ni Alloys

Non-equilibrium solidification structures of Cu55Ni45 and Cu55Ni43Co2 alloys were prepared by the molten glass purification cycle superheating method.The variation of the recalescence phenomenon with the degree of undercooling in the rapid solidification process was investigated using an infrared thermometer.The addition of the Co element affected the evolution of the recalescence phenomenon in Cu-Ni alloys.The images of the solid-liquid interface migration during the rapid solidification of supercooled melts were captured by using a high-speed camera.The solidification rate of Cu-Ni alloys,with the addition of Co elements,was explored.Finally,the grain refinement structure with low supercooling was characterised using electron backscatter diffraction(EBSD).The effect of Co on the microstructural evolution during nonequilibrium solidification of Cu-Ni alloys under conditions of small supercooling is investigated by comparing the microstructures of Cu55Ni45 and Cu55Ni43Co2 alloys.The experimental results show that the addition of a small amount of Co weakens the recalescence behaviour of the Cu55Ni45 alloy and significantly reduces the thermal strain in the rapid solidification phase.In the rapid solidification phase,the thermal strain is greatly reduced,and there is a significant increase in the characteristic undercooling degree.Furthermore,the addition of Co and the reduction of Cu not only result in a lower solidification rate of the alloy,but also contribute to the homogenisation of the grain size.

An efficient method for constructing a random insertional mutant library for forward genetics in Nannochloropsis oceanica

Insertional mutation,phenotypic evaluation,and mutated gene cloning are widely used to clone genes from scratch.Exogenous genes can be integrated into the genome during non-homologous end joining(NHEJ)of the double-strand breaks of DNA,causing insertional mutation.The random insertional mutant library constructed using this method has become a method of forward genetics for gene cloning.However,the establishment of a random insertional mutant library requires a high transformation efficiency of exogenous genes.Many microalgal species show a low transformation efficiency,making constructing random insertional mutant libraries difficult.In this study,we established a highly efficient transformation method for constructing a random insertional mutant library of Nannochloropsis oceanica,and tentatively tried to isolate its genes to prove the feasibility of the method.A gene that may control the growth rate and cell size was identified.This method will facilitate the genetic studies of N.oceanica,which should also be a reference for other microalgal species.

Instance Segmentation of Characters Recognized in Palmyrene Aramaic Inscriptions

This study presents a single-class and multi-class instance segmentation approach applied to ancient Palmyrene inscriptions,employing two state-of-the-art deep learning algorithms,namely YOLOv8 and Roboflow 3.0.The goal is to contribute to the preservation and understanding of historical texts,showcasing the potential of modern deep learning methods in archaeological research.Our research culminates in several key findings and scientific contributions.We comprehensively compare the performance of YOLOv8 and Roboflow 3.0 in the context of Palmyrene character segmentation—this comparative analysis mainly focuses on the strengths and weaknesses of each algorithm in this context.We also created and annotated an extensive dataset of Palmyrene inscriptions,a crucial resource for further research in the field.The dataset serves for training and evaluating the segmentation models.We employ comparative evaluation metrics to quantitatively assess the segmentation results,ensuring the reliability and reproducibility of our findings and we present custom visualization tools for predicted segmentation masks.Our study advances the state of the art in semi-automatic reading of Palmyrene inscriptions and establishes a benchmark for future research.The availability of the Palmyrene dataset and the insights into algorithm performance contribute to the broader understanding of historical text analysis.

A Method for Detecting and Recognizing Yi Character Based on Deep Learning

Aiming at the challenges associated with the absence of a labeled dataset for Yi characters and the complexity of Yi character detection and recognition,we present a deep learning-based approach for Yi character detection and recognition.In the detection stage,an improved Differentiable Binarization Network(DBNet)framework is introduced to detect Yi characters,in which the Omni-dimensional Dynamic Convolution(ODConv)is combined with the ResNet-18 feature extraction module to obtain multi-dimensional complementary features,thereby improving the accuracy of Yi character detection.Then,the feature pyramid network fusion module is used to further extract Yi character image features,improving target recognition at different scales.Further,the previously generated feature map is passed through a head network to produce two maps:a probability map and an adaptive threshold map of the same size as the original map.These maps are then subjected to a differentiable binarization process,resulting in an approximate binarization map.This map helps to identify the boundaries of the text boxes.Finally,the text detection box is generated after the post-processing stage.In the recognition stage,an improved lightweight MobileNetV3 framework is used to recognize the detect character regions,where the original Squeeze-and-Excitation(SE)block is replaced by the efficient Shuffle Attention(SA)that integrates spatial and channel attention,improving the accuracy of Yi characters recognition.Meanwhile,the use of depth separable convolution and reversible residual structure can reduce the number of parameters and computation of the model,so that the model can better understand the contextual information and improve the accuracy of text recognition.The experimental results illustrate that the proposed method achieves good results in detecting and recognizing Yi characters,with detection and recognition accuracy rates of 97.5%and 96.8%,respectively.And also,we have compared the detection and recognition algorithms proposed in this paper with other typical algorithms.In these comparisons,the proposed model achieves better detection and recognition results with a certain reliability.

Fireworks Optimization with Deep Learning-Based Arabic Handwritten Characters RecognitionModel

Handwritten character recognition becomes one of the challenging research matters.More studies were presented for recognizing letters of various languages.The availability of Arabic handwritten characters databases was confined.Almost a quarter of a billion people worldwide write and speak Arabic.More historical books and files indicate a vital data set for many Arab nationswritten in Arabic.Recently,Arabic handwritten character recognition(AHCR)has grabbed the attention and has become a difficult topic for pattern recognition and computer vision(CV).Therefore,this study develops fireworks optimizationwith the deep learning-based AHCR(FWODL-AHCR)technique.Themajor intention of the FWODL-AHCR technique is to recognize the distinct handwritten characters in the Arabic language.It initially pre-processes the handwritten images to improve their quality of them.Then,the RetinaNet-based deep convolutional neural network is applied as a feature extractor to produce feature vectors.Next,the deep echo state network(DESN)model is utilized to classify handwritten characters.Finally,the FWO algorithm is exploited as a hyperparameter tuning strategy to boost recognition performance.Various simulations in series were performed to exhibit the enhanced performance of the FWODL-AHCR technique.The comparison study portrayed the supremacy of the FWODL-AHCR technique over other approaches,with 99.91%and 98.94%on Hijja and AHCD datasets,respectively.

Audio2AB:Audio-driven collaborative generation of virtual character animation

Background Considerable research has been conducted in the areas of audio-driven virtual character gestures and facial animation with some degree of success.However,few methods exist for generating full-body animations,and the portability of virtual character gestures and facial animations has not received sufficient attention.Methods Therefore,we propose a deep-learning-based audio-to-animation-and-blendshape(Audio2AB)network that generates gesture animations and ARK it's 52 facial expression parameter blendshape weights based on audio,audio-corresponding text,emotion labels,and semantic relevance labels to generate parametric data for full-body animations.This parameterization method can be used to drive full-body animations of virtual characters and improve their portability.In the experiment,we first downsampled the gesture and facial data to achieve the same temporal resolution for the input,output,and facial data.The Audio2AB network then encoded the audio,audio-corresponding text,emotion labels,and semantic relevance labels,and then fused the text,emotion labels,and semantic relevance labels into the audio to obtain better audio features.Finally,we established links between the body,gestures,and facial decoders and generated the corresponding animation sequences through our proposed GAN-GF loss function.Results By using audio,audio-corresponding text,and emotional and semantic relevance labels as input,the trained Audio2AB network could generate gesture animation data containing blendshape weights.Therefore,different 3D virtual character animations could be created through parameterization.Conclusions The experimental results showed that the proposed method could generate significant gestures and facial animations.

A Novel 6G Scalable Blockchain Clustering-Based Computer Vision Character Detection for Mobile Images

6G is envisioned as the next generation of wireless communication technology,promising unprecedented data speeds,ultra-low Latency,and ubiquitous Connectivity.In tandem with these advancements,blockchain technology is leveraged to enhance computer vision applications’security,trustworthiness,and transparency.With the widespread use of mobile devices equipped with cameras,the ability to capture and recognize Chinese characters in natural scenes has become increasingly important.Blockchain can facilitate privacy-preserving mechanisms in applications where privacy is paramount,such as facial recognition or personal healthcare monitoring.Users can control their visual data and grant or revoke access as needed.Recognizing Chinese characters from images can provide convenience in various aspects of people’s lives.However,traditional Chinese character text recognition methods often need higher accuracy,leading to recognition failures or incorrect character identification.In contrast,computer vision technologies have significantly improved image recognition accuracy.This paper proposed a Secure end-to-end recognition system(SE2ERS)for Chinese characters in natural scenes based on convolutional neural networks(CNN)using 6G technology.The proposed SE2ERS model uses the Weighted Hyperbolic Curve Cryptograph(WHCC)of the secure data transmission in the 6G network with the blockchain model.The data transmission within the computer vision system,with a 6G gradient directional histogram(GDH),is employed for character estimation.With the deployment of WHCC and GDH in the constructed SE2ERS model,secure communication is achieved for the data transmission with the 6G network.The proposed SE2ERS compares the performance of traditional Chinese text recognition methods and data transmission environment with 6G communication.Experimental results demonstrate that SE2ERS achieves an average recognition accuracy of 88%for simple Chinese characters,compared to 81.2%with traditional methods.For complex Chinese characters,the average recognition accuracy improves to 84.4%with our system,compared to 72.8%with traditional methods.Additionally,deploying the WHCC model improves data security with the increased data encryption rate complexity of∼12&higher than the traditional techniques.

Frontier Progress of Landscape Character Assessment in China

The role of Landscape Character Assessment(LCA)at the level of territorial landscape governance spans both natural and social sciences.By analyzing the development history,research distribution,methods and applications of cutting-edge cases of LCA in China,the following conclusions are drawn:①the LCA research in China originated earlier than that in Europe,but has not yet been systematically applied to the implementation of urban and rural planning at all levels;②the fundamental theory of LCA in China has been well constructed,with three main research directions:technologyled,assessment-led,and assessment combined with other theories;③the development of LCA in rural areas is more mature than in urban areas,but the progress of research is uneven across regions;④the current research presents significant“bottom-up”academic characteristics,and there is an urgent need for government decision-making authorities and academia to jointly promote a“top-down”standardized governance mechanism to comprehensively promote the modernization of territorial landscape governance.

Progress in the evaluation tool of Character Strengths and its application in nursing staff

Character Strengths is a group of positive personality traits reflected through cognition,behavior,and emotion,which play a positive role in improving happiness,alleviating negative emotions,and maintaining physical and mental health.This article reviews the concept,content,methods,evaluation tools,and application progress of Character Strengths in nurses,so as to provide a reference for clinical managers and improve the quality of life,mental health,and professional satisfaction of n urses.

Chip Surface Character Recognition Based on OpenCV

Chip surface character recognition is an important part of quality inspection in the field of microelectronics manufacturing.By recognizing the character information on the chip,automated production,quality control,and data collection and analysis can be achieved.This article studies a chip surface character recognition method based on the OpenCV vision library.Firstly,the obtained chip images are preprocessed.Secondly,the template matching method is used to locate the chip position.In addition,the surface characters on the chip are individually segmented,and each character image is extracted separately.Finally,a Support Vector Machine(SVM)is used to classify and recognize characters.The results show that this method can accurately recognize the surface characters of chips and meet the requirements of chip quality inspection.

Roles of Mutant TP53 Gene in Cancer Development and Progression

TP53 is a tumor suppressor gene that is mutated in most cancer types and has been extensively studied in cancer research.p53 plays a critical role in regulating the expression of target genes and is involved in key processes such as apoptosis,cell cycle regulation,and genomic stability,earning it the title“guardian of the genome.”Numerous studies have demonstrated p53’s influence on and regulation of autophagy,ferroptosis,the tumor microenvironment,and cell metabolism,all of which contribute to tumor suppression.Alterations in p53,specifically mutant p53(mutp53),not only impair its tumor-suppressing functions but also enhance oncogenic characteristics.Recent data indicate that mutp53 is strongly associated with poor prognosis and advanced cancers,making it an ideal target for the development of novel cancer therapies.This review summarizes the post-translational modifications of p53,the mechanisms of mutp53 accumulation,and its gain-of-function,based on previous findings.Additionally,this review discusses its impact on metabolic homeostasis,ferroptosis,genomic instability,the tumor microenvironment,and cancer stem cells,and highlights recent advancements in mutp53 research.

Physiological Character and Gene Mapping in a New Green-revertible Albino Mutant in Rice

A green-revertible albino mutant-Qiufeng M was found from the japonica rice （Oryza sativa L. ssp. japonica） Qiufeng in the field. The first three leaves of the mutant were albino with some green. The leaf color became pale green since the fourth leaf and the glume had the same phenomenon as the first three leaves. The measuring data of the pigment content confirmed the visually observed results. It truly had a remarkable changing process in the leaf color in Qiufeng M. Comparison of the main agronomic characters between Qiufeng and Qiufeng M indicated that the neck length and grain weight showed significant difference at the 1% level, and other characters were not different. Genetic analysis showed that the green-revertible albino trait was controlled by a single recessive nucleic gene. Using 209 recessive mutant individuals in the F2 population derived from the cross Pei＇ai 64S × Qiufeng M, a gene, tentatively named gra（t）, was located between the SSR markers of RM475 and RM2-22 on the long arm of chromosome 2. The genetic distance were 17.3 cM and 2.9 cM respectively.

Grain Filling Characteristics of Giant Embryo Rice Mutant

[Objective] The aim was to explore the relevance between decreased 1 000-grain weight and grain-filling characteristics of rice giant embryo mutant.[Method] Richards equation was used to describe grain filling processes of giant embryo mutant MH-ge1 and its corresponding wild type so as to analyze the parameters of grain-filling characteristics of materials.[Result] The initial filling power of mutant MH-ge1 was higher than the wild type MH86,however,the maximum filling rate,the mean filling rate and the final increment of mutant MH-ge1 was lower than those of MH86.[Conclusion] The grain-filling characteristics of mutant MH-ge1 were poorer than the wild type MH86,thus resulting in the decrease of the grain weight of the giant embryo rice.

Characterization and mapping of a novel light-dependent lesion mimic mutant lmm6 in rice(Oryza sativa L.)

A novel rice lesion mimic mutant (LMM) was isolated from an ethane methyl sulfonate (EMS)-induced 02428 mutant bank. The mutant, tentatively designated as lmm6, develops necrotic lesions in the whole growth period along with changes in several important agronomic traits. We found that the initiation of the lesions was induced by light and cel death occurred in lmm6 accompanied with accumulation of reactive oxygen species (ROS). The lower chlorophyl content, soluble protein content and superoxide dismutase (SOD) activity, the higher malondialdehyde (MDA) content were detected in lmm6 than in the wild type (WT). Moreover, the observation by transmission electronic microscope (TEM) demonstrated that some organel es were damaged and the stroma lamel a of chloroplast was irregular and loose in mesophyl cel of lmm6. In addition, lmm6 was more resistant than WT to rice blast fungus Magnaporthe grisea infection, which was consistent with increased expression of four genes involved in the defense-related reaction. Genetic analysis showed that mutant trait of lmm6 is inherited as a monogenic recessive nuclear gene located on the long arm of chromosome 6. Using simple sequence repeat (SSR) markers, the target gene was ifnal y delimited to an interval of 80.8 kb between markers MM2359 and MM2370, containing 7 annotated genes. Taken together, our results provide the information to identify a new gene involved in rice lesion mimic, which wil be helpful in clarifying the mechanism of cel death and disease resistance in rice.

Characterization and mapping of QTLs on chromosome 2D for grain size and yield traits using a mutant line induced by EMS in wheat

Production of mutants with altered phenotypes is a powerful approach for determining the biological functions of genes in an organism. In this study, a high-grain-weight mutant line M8008 was identified from a library of mutants of the common wheat cultivar YN15 treated with ethylmethane sulfonate(EMS). F2 and F2:3generations produced from crosses of M8008 × YN15(MY) and M8008 × SJZ54(MS) were used for genetic analysis. There were significant differences between M8008 and YN15 in plant height(PH), spike length(SL),fertile spikelet number per spike(FSS), grain width(GW), grain length(GL), GL/GW ratio(GLW), and thousand-grain weight(TGW). Most simple correlation coefficients were significant for the investigated traits, suggesting that the correlative mutations occurred in M8008. Approximately 21% of simple sequence repeat(SSR) markers showed polymorphisms between M8008 and YN15, indicating that EMS can induce a large number of mutated loci. Twelve quantitative trait loci(QTLs) forming QTL clusters(one in MY and two in MS) were detected. The QTL clusters coinciding with(MY population) or near(MS population) the marker wmc41 were associated mainly with grain-size traits, among which the M8008 locus led to decreases in GW, factor form density(FFD), and TGW and to increases in GLW. The cluster in the wmc25–barc168 interval in the MS population was associated with yield traits, for which the M8008 locus led to decreased PH, spike number per plant(SN), and SL.

Plant Hormone Resistance and Agronomic Characteristics of the MT10 Mutant in Rice

The MT10 mutant plants had resistances to auxin.Under light and dark culture,the roots of MT10 seedlings had shown less lateral roots and short lateral roots.In soil,MT10 seedlings had shown not only no changed agronomic characteristics but also no significant difference with WT.

Physiological and molecular characteristics of two ploidy mutants in Myrica rubracv.Dongkui

In this study, two ploidy mutant lines ofMyrica rubra cv. Dongkui （DK） were identiifed and named as DB1 and DB2. The lforal organ, leaf cel structure, ploidy, and number of chromosomes of the two mutants were investigated. Meanwhile, anthocyanin contents at different developmental stages were analyzed, and the Cy-3-glu contents of DB1 and DB2 at the ful ripe stages are signiifcantly higher than that of DK by 27.84 and 23.51%, respectively. Furthermore, 6 RNA libraries at two developmental stages （young fruit stage and ful ripe stage） were built for RNA-Seq. By mapping to the reference database, 28407, 28043, and 28683 genes were detected in the young fruit of DB1, DB2, and DK, respectively, while 28040, 22256, and 27351 genes were detected in the ful ripe stage, respectively. There were 281 differentialy expressed genes between DB1 and DK, with 123 and 158 genes up-regulated and down-regulated, respectively, and 47 differentialy expressed genes between DB2 and DK, of which 8 and 39 genes were up-regulated and down-regulated. Using real-time PCR, the expression levels of the eight functional genes at different developmental stages of the fruit were also analyzed. These comprehensive analyses showed that both mutants are different from DK, which is the result of natural doubling of ploidy, thereby generating a pleiotropic effect. As we known, it is the ifrst report to study the relationship between bayberry ploidy alterations and genes involved in regulation of fruit mutations, which wil help to identify the morphological and cytological characteristics ofM. rubragermplasm, and provide a theoretical basis and technical support for genetic improvement and creation of breeding resources.

Characterization and mapping of a new male sterility mutant of anther advanced dehiscence(t) in rice

Anther dehiscence is very important for pollen maturation and release. The mutants of anther dehiscence in rice （Oryza sativa L.） are few, and related research remains poor. A male sterility mutant of anther dehiscence in advance, add（t）, has been found in Minghui 63 and its sterility is not sensitive to thermo-photo. To learn the character of sterilization and the function of the add（t） gene, the morphological and cytological studies on the anther and pollen, the ability of the pistil being fertilized, inheritance of the mutant, and mapping of add（t） gene have been conducted. The anther size is normal but the color is white in the mutant against the natural yellow in the wild-type. The pollen is malformed, unstained, and small in the KI-I2 solution. The anther dehiscence is in advance at the bicellular pollen stage. A crossing test indicated that the grain setting ratio of the add（t） is significantly lower than that of the CMS line 2085A. The ability of the pistil being fertilized is most probably decreased by the add（t） gene. The male sterility is controlled by a single recessive gene of add（t）. This gene is mapped between the markers of R02004 （InDel） and RM300 （SSR） on chromosome 2, and the genetic distance from the add（t） gene to these markers is 0.78 cM and 4.66 cM, respectively.