Recognition Step Research Articles

Hypergraph Neural Network for Emotion Recognition in Conversations

Modeling conversational context is an essential step for emotion recognition in conversations. Existing works still suffer from insufficient utilization of local context information and remote context information. This article designs a hypergraph neural network, namely HNN-ERC, to better utilize local and remote contextual information. HNN-ERC combines the recurrent neural network with the conventional hypergraph neural network to strengthen connections between utterances and make each utterance receive information from other utterances better. The proposed model has empirically achieved state-of-the-art results on three benchmark datasets, demonstrating the effectiveness and superiority of the new model.

Elf1 promotes Rad26's interaction with lesion-arrested Pol II for transcription-coupled repair.

Transcription-coupled nucleotide excision repair (TC-NER) is a highly conserved DNA repair pathway that removes bulky lesions in the transcribed genome. Cockayne syndrome B protein (CSB), or its yeast ortholog Rad26, has been known for decades to play important roles in the lesion-recognition steps of TC-NER. Another conserved protein ELOF1, or its yeast ortholog Elf1, was recently identified as a core transcription-coupled repair factor. How Rad26 distinguishes between RNA polymerase II (Pol II) stalled at a DNA lesion or other obstacles and what role Elf1 plays in this process remains unknown. Here, we present cryo-EM structures of Pol II-Rad26 complexes stalled at different obstacles that show that Rad26 uses a common mechanism to recognize a stalled Pol II, with additional interactions when Pol II is arrested at a lesion. A cryo-EM structure of lesion-arrested Pol II-Rad26 bound to Elf1 revealed that Elf1 induces further interactions between Rad26 and a lesion-arrested Pol II. Biochemical and genetic data support the importance of the interplay between Elf1 and Rad26 in TC-NER initiation. Together, our results provide important mechanistic insights into how two conserved transcription-coupled repair factors, Rad26/CSB and Elf1/ELOF1, work together at the initial lesion recognition steps of transcription-coupled repair.

Research and Application of Knowledge Graph Design for Interactive Teaching Platform Based on Artificial Intelligence

Abstract With the rapid development of education digitization, artificial intelligence (AI) and knowledge graph, their application in the field of education has received extensive attention and research by many scholars. In this paper, the design of knowledge graphs for AI teaching platforms is studied. Based on the construction process of a general knowledge graph, teaching resources’ text is processed through the steps of knowledge entity recognition and relationship extraction, which is applied to the intelligent interactive teaching platform. On this basis, the article designs a method for teaching resource retrieval based on semantic analysis to provide a path for further teaching applications of knowledge graphs. After completing its design, application and testing are carried out to test its feasibility and effectiveness. The score of the experimental class that used the knowledge graph increased by 17.68, which is about 22.47% year-on-year compared to the pre-class period; the score of the control class also rose, with a rise of 9.41, which is about 11.53% year-on-year. The results indicate that a knowledge graph-equipped teaching platform can better meet the needs of interactive teaching and enhance students’ learning abilities.

Expanding limits of artificial enzymes: unprecedented catalysis by an oxidase nanozyme in activating a structural protein for covalent crosslinking and conferring remarkable proteolytic resistance.

Nature has endowed us with some complex enzymes capable of utilizing proteins as their substrates to generate functional proteins through post-translational modification. However, nanozymes' interplay with proteins as substrates is scarce, with their chemistry predominantly established using only small molecule substrates, featuring a significant gap in this area. Due to the huge prospects of nanozymes in biotechnological and therapeutic interventions, studies establishing the unexplored roles of nanozymes in the biological environments and their interplay beyond small molecule substrates warrant immediate attention. In this study, we unveil the unprecedented role of a Mn-based oxidase nanozyme (MnN) in activating a structural protein, collagen, and covalently crosslinking its tyrosine residues with only a trace amount of tannic acid (TA) without compromising its triple-helical structural integrity. While therapeutic applications demand materials prepared from collagen, the current chemical and physical crosslinking of collagen often presents significant challenges such as toxicity, denaturation, or high costs. MnN lucidly accomplishes crosslinking interplay at its 101 facets using oxygen as a co-substrate under mild conditions. This process takes advantage of MnN being active at mild acidic pH where collagen preferentially exists as a soluble triple helix (monomeric form), exposing functionalities and enhancing the crosslinking degree. Importantly, this reaction also confers 100% resistance to collagenase attack on the collagen tendon-derived biological material. The catalyzed TA-tyrosine linkage in the telopeptide region of collagen probably impedes the initial recognition step of collagenase, providing robust protection against its degradative action. Our study not only expands the repertoire of nanozymes' substrates beyond the existing library of small molecules but also establishes a significant step toward designing a gold standard for collagen crosslinking. With biomedical applications demanding biomaterials derived from protein scaffolds with preserved structural integrity, our investigation bridges the gap between nanozymes' chemistry and crosslinking proteins, opening exciting prospects for biomaterial development.

Revolutionizing Sales with Phone Systems: Design and Implementation Best Practices

The technique of classifying a face as recognized or unknown in computer vision is known as face recognition. The procedure follows a pipeline that includes steps for data collection, detection, pre-processing, and recognition. This research focuses on with the assumption that photos have already been processed at the end of the pipeline pre-processed and gathered Face recognition software in the past has relied on the full face. Their algorithms take a face picture as input. We propose a new method in which the individual facial segments, such as the left, are sent into the recognition algorithm. The four components that make up the face are the left eye, the right eye, the nose, and the mouth. After that, the output of each algorithm is labelled, and it is categorized as belonging to right feature class. According to the findings of our research, the granularity of information gathered for each subdivision class has the potential to be utilized in order to create a higher accuracy rate than can be achieved by using the complete face technique.

A Transient π-π or Cation-π Interaction between Degron and Degrader Dual Residues: A Key Step for the Substrate Recognition and Discrimination in the Processive Degradation of SulA by ClpYQ (HslUV) Protease in Escherichia coli.

The Escherichia coli ATP-dependent ClpYQ protease constitutes ClpY ATPase/unfoldase and ClpQ peptidase. The Tyr91st residue within the central pore-I site of ClpY-hexamer is important for unfolding and translocating substrates into the catalytic site of ClpQ. We have identified the degron site (GFIMRP147th) of SulA, a cell-division inhibitor recognized by ClpYQ and that the Phe143rd residue in degron site is necessary for SulA native folded structure. However, the functional association of this degron site with the ClpYQ degrader is unknown. Here, we investigated the molecular insights into substrate recognition and discrimination by the ClpYQ protease. We found that the point mutants ClpYY91FQ, ClpYY91HQ, and ClpYY91WQ, carrying a ring structure at the 91st residue of ClpY, efficiently degraded their natural substrates, evidenced by the suppressed bacterial methyl-methane-sulfonate (MMS) sensitivity, the reduced β-galactosidase activity of cpsB::lacZ, and the lowest amounts of MBP-SulA in both in vivo and in vitro degradation analyses. Alternatively, mimicking the wild-type SulA, SulAF143H, SulAF143K and SulAF143W, harboring a ring structure or a cation side-group in 143rd residue of SulA, were efficiently degraded by ClpYQ in the bacterial cells, also revealing shorter half-lives at 41 °C and higher binding affinities towards ClpY in pull-down assays. Finally, ClpYY91FQ and ClpYY91HQ, were capable of effectively degrading SulAF143H and SulAF143K, highlighting a correspondingly functional interaction between the SulA 143rd and ClpY 91st residues. According to the interchangeable substituted amino acids, our results uniquely indicate that a transient π-π or cation-π interaction between the SulA 143rd and ClpY 91st residues could be aptly gripped between the degron site of substrates and the pore site of proteases (degraders) for substrate recognition and discrimination of the processive degradation.

Hybrid Arabic handwritten character segmentation using CNN and graph theory algorithm

Arabic handwritten words segmentation is considered a challenging step in Arabic optical character recognition. In this work, we propose a convolutional neural network-based segmentation of handwritten Arabic words. After generating candidate segmentation points, we applied a graph theory-based technique to divide the word into sub-words. Our approach reached a 96% correct sub-word segmentation rate. Additionally, we develop a Segmentation Hypothesis Graph (SHG) to generate candidate characters for each sub-word. We employ both a manually designed convolutional neural network and a transfer learning-based CNN using the pretrained AlexNet model. The accuracy, 96.97%, of the pretrained model outperformed the manually and state of the art models trained on the same dataset. Finally, we used recognition confidence to validate the segmentation points. Hence, our approach correctly segments 88% of the words into characters and successfully resolves 100% of the overlapping cases.

Insights into the ISG15 transfer cascade by the UBE1L activating enzyme

The attachment of the ubiquitin-like protein ISG15 to substrates by specific E1-E2-E3 enzymes is a well-established signalling mechanism of the innate immune response. Here, we present a 3.45 Å cryo-EM structure of a chemically trapped UBE1L-UBE2L6 complex bound to activated ISG15. This structure reveals the details of the first steps of ISG15 recognition and UBE2L6 recruitment by UBE1L (also known as UBA7). Taking advantage of viral effector proteins from severe acute respiratory coronavirus 2 (SARS-CoV-2) and influenza B virus (IBV), we validate the structure and confirm the importance of the ISG15 C-terminal ubiquitin-like domain in the adenylation reaction. Moreover, biochemical characterization of the UBE1L-ISG15 and UBE1L-UBE2L6 interactions enables the design of ISG15 and UBE2L6 mutants with altered selectively for the ISG15 and ubiquitin conjugation pathways. Together, our study helps to define the molecular basis of these interactions and the specificity determinants that ensure the fidelity of ISG15 signalling during the antiviral response.

A multitasking multi-objective differential evolution gene selection algorithm enhanced with new elite and guidance strategies for tumor identification

A key preprocessing step in tumor recognition based on microarray expression profile data and machine learning is to identify tumor marker genes. Gene selection aims to select the most relevant gene subset from the original ultra-high dimensional microarray expression profile data to improve tumor identification performance. Inspired by evolutionary multitasking (EMT) and multi-objective optimization, this paper puts forward a novel multitasking multi-objective differential evolution gene selection algorithm (MMODE) which uses new elite and guidance strategies to select the best gene subsets. MMODE initializes two different populations according to different filtering criteria to increase the diversity of the search. These two populations guide their respective populations to search in the optimal direction through knowledge transfer in the evolutionary process. In addition, MMODE employs new elite and guidance strategies that enables individuals to narrow the search range and jump out of local optima. The proposed algorithm is validated on 13 publicly available microarray expression datasets in comparison with state-of-the-art gene selection algorithms. The experimental results show that MMODE can find smaller gene subsets and achieve higher classification accuracy compared with other algorithms.

Molecular Mechanism of Labelling Functional Cysteines by Heterocyclic Thiones.

Heterocyclic thiones have recently been identified as reversible covalent warheads, consistent with their mild electrophilic nature. Little is known so far about their mechanism of action in labelling nucleophilic sidechains, especially cysteines. The vast number of tractable cysteines promotes a wide range of target proteins to examine; however, our focus was put on functional cysteines. We chose the main protease of SARS-CoV-2 harboring Cys145 at the active site that is a structurally characterized and clinically validated target of covalent inhibitors. We screened an in-house, cysteine-targeting covalent inhibitor library which resulted in several covalent fragment hits with benzoxazole, benzothiazole and benzimidazole cores. Thione derivatives and Michael acceptors were selected for further investigations with the objective of exploring the mechanism of inhibition of the thiones and using the thoroughly characterized Michael acceptors for benchmarking our studies. Classical and hybrid quantum mechanical/molecular mechanical (QM/MM) molecular dynamics simulations were carried out that revealed a new mechanism of covalent cysteine labelling by thione derivatives, which was supported by QM and free energy calculations and by a wide range of experimental results. Our study shows that the molecular recognition step plays a crucial role in the overall binding of both sets of molecules.

Air Target Attack Intention Recognition Analysis

Aimed at the problem of air target attack intention recognition, based on the analysis of the influencing factors of air target attack intention, the common methods of air target attack intention recognition were analyzed , the air target attack intention recognition method based on fuzzy similarity was proposed, the basic steps of air target attack intention recognition based on fuzzy similarity were proposed, and the establishment method of membership function was analyzed.

VGG-based discrimination for pneumonia images enhanced by Generative Adversarial Network

Due to problems such as blurring of chest X-ray image and double shadow of heart and lung, the accuracy of judging whether there is COVID-19 is not high. This study intends to use Generative Adversarial Network (GAN) to segment images, and then put the segmented image set into the classification algorithm to determine whether it is infected by the pneumonia. The data set is divided into “normal” and “pneumonia”. In order to remove the interference of rib and heart on lung contour in X-ray image, GAN technology is used to segment the original data set. The segmented image only retains the most important lung silhouette, which is convenient for the next step of recognition and information extraction. The classification algorithm uses Visual Geometry Group (VGG)-16. During the training, only fully connected layer 1 and fully connected layer 2 layers were trained, and the rest layers were frozen. When the model loads the original weight, the top level of the model will be replaced by the head model. The head model is placed on the basic model and becomes part of the actual training model to determine the best weight. After the original X-ray image is processed by GAN, the accuracy of the segmented lung image is improved from 84% to 98%. Loss decreased from 17% to 3%, and the accuracy of classification results was greatly improved. Therefore, it is proved that the segmented image can effectively remove irrelevant parts in the recognition process, and the accuracy is improved due to the removal of noise interference and clearer lung contour.

Application of dynamic time warping optimization algorithm in speech recognition of machine translation

Speech recognition is the foundation of human-computer interaction technology and an important aspect of speech signal processing, with broad application prospects. Therefore, it is very necessary to recognize speech. At present, speech recognition has problems such as low recognition rate, slow recognition speed, and severe interference from other factors. This paper studied speech recognition based on dynamic time warping (DTW) algorithm. By introducing speech recognition, the specific steps of speech recognition were understood. Before performing speech recognition, the speech that needs to be recognized needs to be converted into a speech sequence using an acoustic model. Then, the DTW algorithm was used to preprocess speech recognition, mainly by sampling and windowing the speech. After preprocessing, speech feature extraction was carried out. After feature extraction was completed, speech recognition was carried out. Through experiments, it can be found that the recognition rate of speech recognition on the basis of DTW algorithm was very high. In a quiet environment, the recognition rate was above 93.85 %, and the average recognition rate of the 10 selected testers was 95.8 %. In a noisy environment, the recognition rate was above 91.4 %, and the average recognition rate of the 10 selected testers was 93 %. In addition to high recognition rate, DTW based speech recognition also had a very fast speed for vocabulary recognition. Based on the DTW algorithm, speech recognition not only has a high recognition rate, but also has a faster recognition speed.

Study the effect of eye diseases on the performance of iris segmentation and recognition using transfer deep learning methods

A new deep learning-based iris recognition system is presented in the current study in the case of eye disease. Current state of art iris segmentation is either based on traditional low accuracy algorithms or heavy-weight deep-based models. In the current study segmentation section, a new iris segmentation method based on illumination correction and a modified circular Hough transform is proposed. The current method also performs a post-processing step to minimize the false positives. Besides, a ground truth of iris images is constructed to evaluate the segmentation accuracy. Many deep learning models (GoogleNet, Inception_ResNet, XceptionNet, EfficientNet, and ResNet50) are applied through the recognition step using the transfer learning approach. In the experiment part, two eye disease-based datasets are used. 684 iris images of individuals with multiple ocular diseases from the Warsaw BioBase V1 and 1,793 iris images from the Warsaw BioBase V2 are also used. The CASIA V3 Interval Iris dataset, which contains 2,639 photographs of healthy iris, is used to train deep models once, and then the transfer learning of this normal-based eye dataset is used to retrain the same deep models using Warsaw BioBase datasets. Different scenarios for training and evaluating participants are used during experiments. The trained models are evaluated using validation accuracy, training time, TPR, FNR, PPR, FDR, and test accuracy. The best accuracies are 98.5% and 97.26%, which are recorded by the ResNet50 (2-layer of transfer learning) model trained on Warsaw BioBase V1 and V2, respectively. Results indicate that the effect of eye diseases is concentrated on the segmentation phase. For recognition, no significant impact is recognized. Some disease that affects the structure (bloody eyes, trauma, iris pigment) can affect the iris recognition step partially. Our study is compared with similar studies in the case of eye diseases. The comparison proves the efficiency and high performance of the proposed methodology against all previous models on the same iris datasets.

MHG-ERC: Multi-hypergraph Feature Aggregation Network for Emotion Recognition in Conversations

The modeling of conversational context is an essential step in Emotion Recognition in Conversations (ERC). To maintain high performance and a low GPU memory consumption, this article proposes a new idea of using multiple hypergraphs to model the conversational context and designs a multi-hypergraph feature aggregation network for ERC. We use context window, speaker information, position information between utterances, and specific step size to construct different hyperedges. Then, various hypergraphs generated by different hyperedges are used to aggregate local and remote context information in turn. Experiments on two dialogue emotion datasets, IEMOCAP and MELD, demonstrate the effectiveness and superiority of this new model. In addition, our model requires only relatively low GPU memory consumption.

AHR-mediated DNA damage contributes to BaP-induced cardiac malformations in zebrafish

Benzo[a]pyrene (BaP) is a representative polycyclic aromatic hydrocarbon widely present in the environment. We previously reported that the aryl hydrocarbon receptor (AHR) mediates BaP-induced apoptosis and cardiac malformations in zebrafish embryos, but the underlying molecular mechanisms were unclear. Since BaP is a mutagenetic compound, we hypothesize that BaP induces apoptosis and heart defects via AHR-mediated DNA damage. In this study, zebrafish embryos were exposed to BaP at a concentration of 0.1 μM from 2 to 72 h post fertilization, either with or without inhibitors/agonists. AHR activity and levels of reactive oxygen species (ROS) were examined under a fluorescence microscope. mRNA expression levels were quantified by qPCR. DNA damage and apoptosis were detected by immunofluorescence. Our findings revealed that BaP exposure significantly increased BPDE-DNA adducts, mitochondrial damage, apoptosis and heart defects in zebrafish embryos. These effects were counteracted by inhibiting AHR/cyp1a1 using pharmaceutical inhibitors or genetic knockdown. Furthermore, we observed that spironolactone, an antagonist of nucleotide excision repair (NER), significantly enhanced BaP-induced BPDE-DNA adducts, mitochondrial damage, apoptosis and heart malformation rates. Conversely, SRT1720, a SIRT1 agonist, reduced the adverse effects of BaP. Supplementation with spironolactone also enhanced γ-H2AX signals in the heart of zebrafish embryos exposed to BaP. Additional experiments demonstrated that BaP suppressed the expression of SIRT1. We further established that AHR, when activated by BaP, directly inhibited SIRT1 transcription, leading to downregulation of XPC and XPA, which are essential NER genes involved in the recognition and verification steps of the NER process. Taken together, our results indicate that AHR mediates BaP-induced DNA damage in the heart of zebrafish embryos by inducing BPDE-DNA adduct formation via the AHR/Cyp1a1 signalling pathway, as well as suppressing NER via AHR-mediated inhibition of SIRT1.

Chromosome-level Dinobdella ferox genome provided a molecular model for its specific parasitism

BackgroundDinobdella ferox is the most frequently reported leech species parasitizing the mammalian nasal cavity. However, the molecular mechanism of this special parasitic behavior has remained largely unknown.MethodsPacBio long-read sequencing, next-generation sequencing (NGS), and Hi-C sequencing were employed in this study to generate a novel genome of D. ferox, which was annotated with strong certainty using bioinformatics methods. The phylogenetic and genomic alterations of D. ferox were then studied extensively alongside the genomes of other closely related species. The obligatory parasitism mechanism of D. ferox was investigated using RNA-seq and proteomics data.ResultsPacBio long-read sequencing and NGS yielded an assembly of 228 Mb and contig N50 of 2.16 Mb. Along Hi-C sequencing, 96% of the sequences were anchored to nine linkage groups and a high-quality chromosome-level genome was generated. The completed genome included 19,242 protein-coding genes. For elucidating the molecular mechanism of nasal parasitism, transcriptome data were acquired from the digestive tract and front/rear ends of D. ferox. Examining secretory proteins in D. ferox saliva helped to identify intimate connections between these proteins and membrane proteins in nasal epithelial cells. These interacting proteins played important roles in extracellular matrix (ECM)–receptor interaction, tight junction, focal adhesion, and adherens junction. The interaction between D. ferox and mammalian nasal epithelial cells included three major steps of pattern recognition, mucin connection and breakdown, and repair of ECM. The remodeling of ECM between epithelial cells of the nasal mucosa and epithelial cells of D. ferox may produce a stable adhesion environment for parasitism.ConclusionsOur study represents the first-ever attempt to propose a molecular model for specific parasitism. This molecular model may serve as a practical reference for parasitism models of other species and a theoretical foundation for a molecular process of parasitism.Graphical abstract

GRRS: Accurate and Efficient Neighborhood Rough Set for Feature Selection

Feature selection is an important preprocessing step in data mining and pattern recognition. The neighborhood rough set (NRS) model is a widely-used rough set model for feature selection on continuous data. All currently known NRS models are defined on a distance metric — mostly the Euclidean distance metric — which invalidates the NRS models in scenarios wherein the Euclidean distance is ineffective, for example, while considering differing attribute weights. We first introduce the concept of space division of granular-rectangular, and then construct the neighborhood radius in our method by describing the relationship between child and parent spaces, which avoids the use of a distance metric and reduces the search space for the neighborhood radius. This greatly improves both the accuracy and efficiency of NRS. In addition, the upper and lower approximations of the granular-rectangular rough sets (GRRSs) comprise equivalence classes; this results in better performance of GRRS in knowledge representation compared with the traditional NRS. Experimental results on public benchmark datasets reveal that our method, GRRS, achieves higher accuracy than ten popular and state-of-the-art feature-selection methods, including two NRS algorithms. Moreover, GRRS outperforms the established NRS algorithms regarding efficiency, including the state-of-the-art NRS algorithm, GBNRS. All code has been released as an open libary called GRRS: https://github.com/syxiaa/GRRS.

Artificial Intelligence in Tongue Image Recognition

Tongue image recognition is a traditional Chinese medicine diagnosis method, which uses the shape, color, and texture of the tongue to judge the health of the human body. With the rapid development of artificial intelligence technology, the application of artificial intelligence in the field of tongue recognition has been widely considered. Based on the intelligent analysis of tongue diagnosis in traditional Chinese medicine, this paper reviews the application progress of artificial intelligence in tongue image recognition in recent years and analyzes its potential and challenges in this field. Firstly, this paper introduces three steps of tongue image recognition, including tongue image acquisition, tongue image preprocessing, and tongue image feature analysis. The application of traditional methods and artificial intelligence methods in the whole process of tongue image recognition is reviewed, especially the tongue body segmentation, and the advantages and disadvantages of convolutional neural networks are analyzed and compared. Artificial intelligence can use technologies such as deep learning and computer vision to automatically analyze and extract features from tongue images. By constructing a tongue image recognition model, tongue shape, color, texture, and other features can be accurately recognized and quantitatively analyzed. Finally, this paper summarizes the problems existing in artificial intelligence in tongue image recognition and looks forward to the future developmental direction of this field. It can promote the modernization of TCM diagnostic methods, achieve early disease screening and prevention, personalized medicine and treatment optimization, and support medical research and knowledge accumulation. However, there is still a need for further validation and practice, with a focus on patient privacy and data security.

Micro Expression Spotting through Appearance Based Descriptor and Distance Analysis

Micro-Expressions (MEs) are a typical kind of expressions which are subtle and short lived in nature and reveal the hidden emotion of human beings. Due to processing an entire video, the MEs recognition constitutes huge computational burden and also consumes more time. Hence, MEs spotting is required which locates the exact frames at which the movement of ME persists. Spotting is regarded as a primary step for MEs recognition. This paper proposes a new method for ME spotting which comprises three stages; pre-processing, feature extraction and discrimination. Pre-processing aligns the facial region in every frame based on three landmark points derived from three landmark regions. To do alignment, an in-plane rotation matrix is used which rotates the non-aligned coordinates into aligned coordinates. For feature extraction, two texture based descriptors are deployed; they are Local Binary Pattern (LBP) and Local Mean Binary Pattern (LMBP). Finally at discrimination stage, Feature Difference Analysis is employed through Chi-Squared Distance (CSD) and the distance of each frame is compared with a threshold to spot there frames namely Onset, Apex and Offset. Simulation done over a Standard CASME dataset and performance is verified through Feature Difference and F1-Score. The obtained results prove that the proposed method is superior than the state-of-the-art methods.