Higher Confidence Scores Research Articles

In Silico Analysis of the Molecular Interaction between Anthocyanase, Peroxidase and Polyphenol Oxidase with Anthocyanins Found in Cranberries.

Anthocyanins are bioactive compounds responsible for various physiological processes in plants and provide characteristic colors to fruits and flowers. Their biosynthetic pathway is well understood; however, the enzymatic degradation mechanism is less explored. Anthocyanase (β-glucosidase (BGL)), peroxidase (POD), and polyphenol oxidase (PPO) are enzymes involved in degrading anthocyanins in plants such as petunias, eggplants, and Sicilian oranges. The aim of this work was to investigate the physicochemical interactions between these enzymes and the identified anthocyanins (via UPLC-MS/MS) in cranberry (Vaccinium macrocarpon) through molecular docking to identify the residues likely involved in anthocyanin degradation. Three-dimensional models were constructed using the AlphaFold2 server based on consensus sequences specific to each enzyme. The models with the highest confidence scores (pLDDT) were selected, with BGL, POD, and PPO achieving scores of 87.6, 94.8, and 84.1, respectively. These models were then refined using molecular dynamics for 100 ns. Additionally, UPLC-MS/MS analysis identified various flavonoids in cranberries, including cyanidin, delphinidin, procyanidin B2 and B4, petunidin, pelargonidin, peonidin, and malvidin, providing important experimental data to support the study. Molecular docking simulations revealed the most stable interactions between anthocyanase and the anthocyanins cyanidin 3-arabinoside and cyanidin 3-glucoside, with a favorable ΔG of interaction between -9.3 and -9.2 kcal/mol. This study contributes to proposing a degradation mechanism and seeking inhibitors to prevent fruit discoloration.

Automatic identification of breech face impressions based on deep local features

Breech face impressions are an essential type of physical evidence in forensic investigations. However, their surface morphology is complex and varies based on the machining method used on the gun’s breech face, making traditional handcrafted local feature-based methods exhibit high false rates and are unsuitable for striated impressions. We proposed a deep local feature-based method for firearm identification utilizing Detector-Free Local Feature Matching with Transformers (LoFTR). This method removes the module of feature point detection and directly utilizes self and cross-attention layers in the Transformer to transform the convolved coarse-level feature maps into a series of dense feature descriptors. Subsequently, matches with high confidence scores are filtered based on the score matrix calculated from the dense descriptors. Finally, the screened initial matches are refined into the convolved fine-level features, and a correlation-based approach is used to obtain the exact location of the match. Validation tests were conducted using three authoritative sets of the breech face impressions datasets provided by the National Institute of Standards and Technology (NIST). The validation results show that, compared with the traditional handcrafted local-feature based methods, the proposed method in this paper yields a lower identification error rate. Notably, the method can not only deal with granular impressions, but can also be applied to the striated impressions. The results indicate that the method proposed in this paper can be utilized for comparative analysis of breech face impressions, and provide a new automatic identification method for forensic investigations.

The influence of care home registration type and size on senior care leader’s confidence to provide palliative and end-of-life care: an explanatory sequential mixed methods study

BackgroundCare home staff are key providers of palliative and end-of-life care. Yet, little is known about how care home characteristics can influence care leader’s confidence in their ability to provide optimal palliative and end-of-life care.AimTo understand the influence of care home registration type (nursing, residential or dual registered) and size on senior care leader’s confidence to provide palliative and end-of-life care.DesignAn explanatory sequential mixed methods study comprising an online cross-sectional survey (including the Palliative Care Self-Efficacy Scale) and qualitative individual interviews. Analysis of survey data used a multivariate logistic regression and qualitative interview data used Framework Analysis. A ‘Following the Thread’ method was undertaken for data integration.Setting/participantsUK care home senior care leaders, purposively sampled by registration type, size and geographical location.ResultsThe online survey (N = 107) results indicated that nursing home senior care leaders had higher confidence scores on the Palliative Care Self-Efficacy Scale than residential care home leaders (aOR: 3.85, 95% CI 1.20-12.31, p = 0.02). Care home size did not show effect when adjusting for registration type (medium - aOR 1.71, 95% CI 0.59–4.97, p = 0.33; large – aOR 0.65, 95% CI 0.18–2.30, p = 0.5). Interviews (n = 27) identified three themes that promote confidence, (1) ‘feelings of preparedness’ stemming from staff expertise and experience and care home infrastructure, (2) ‘partnership working’ with external services as a valued member of the multidisciplinary team, and (3) a shared language developed from end-of-life care guidance.ConclusionCare home senior care leader’s confidence is influenced by care home characteristics, particularly availability of on-site registered nurses and the infrastructure of large care homes. All care home leaders benefit from training, working with external, multidisciplinary teams and use of guidance. However, mechanisms to achieve this differed by care home type and size. Further exploration is needed on successful integration of palliative care services and interventions to enhance confidence in residential care homes.

Unveiling the Potential of ChatGPT and YOLOv7 for Evaluating Children's Emotions Using Their Artistic Expressions.

Recent advancements in large language models (LLMs) have sparked considerable interest in their potential applications across various healthcare domains. One promising prospect is leveraging these generative models to accurately predict children's emotions by combining computer vision and natural language processing techniques. However, understanding children's emotional states based on their artistic expressions is equally crucial. To address this challenge, this paper presents a pipelined architecture comprising YOLOv7 and the powerful GPT-3.5 Turbo language model, where YOLOv7 is employed for object detection using art therapy imaging annotations, while GPT-3.5 interprets the sketches. After rigorously evaluating the proposed framework through a series of comprehensive experiments, we observed that our model achieved high confidence scores for both object detection and emotion interpretation. The robust performance of the proposed framework not only aids in explaining children's art but also provides valuable insights for parents and therapists. This capability enables them to better understand children's emotional states based on their artistic expressions, ultimately facilitating improved support and care.

SRPM-ST: Sequential retraining and pseudo-labeling in mini-batches for self-training

An impediment to training accurate classifiers in supervised learning is the scarcity of labeled data. In that respect, semi-supervised learning could help by using both labeled and unlabeled data. A specific form of semi-supervised learning is self-training (ST). In its basic form, ST trains an initial classifier using the labeled data to generate pseudo-labels for the unlabeled set. At this point, either the whole set of pseudo-labeled data or a subset of them with some high confidence scores about the generated pseudo-labels is selected. The selected pseudo-labeled data are then used to update the initial classifier. Although this process can be repeated to generate new pseudo-labels for the unlabeled data, it is typically a tacit assumption up to this point that the classifier is updated once all pseudo-labels are generated—a process to which we refer as the full-batch ST (F-ST) regardless of any confidence score-based subset selection. Here, we show that sequential retraining and pseudo-labeling in mini-batches (SRPM) could potentially improve the performance of the classifier with respect to F-ST. Our empirical results show the existence of a data-dependent mini-batch size for SRPM that is optimal in terms of possessing the least error rate. In practice, this parameter could be treated as a hyperparameter to tune.

Impact of database choice and confidence score on the performance of taxonomic classification using Kraken2

AbstractAccurate taxonomic classification is essential to understanding microbial diversity and function through metagenomic sequencing. However, this task is complicated by the vast variety of microbial genomes and the computational limitations of bioinformatics tools. The aim of this study was to evaluate the impact of reference database selection and confidence score (CS) settings on the performance of Kraken2, a widely used k-mer-based metagenomic classifier. In this study, we generated simulated metagenomic datasets to systematically evaluate how the choice of reference databases, from the compact Minikraken v1 to the expansive nt- and GTDB r202, and different CS (from 0 to 1.0) affect the key performance metrics of Kraken2. These metrics include classification rate, precision, recall, F1 score, and accuracy of true versus calculated bacterial abundance estimation. Our results show that higher CS, which increases the rigor of taxonomic classification by requiring greater k-mer agreement, generally decreases the classification rate. This effect is particularly pronounced for smaller databases such as Minikraken and Standard-16, where no reads could be classified when the CS was above 0.4. In contrast, for larger databases such as Standard, nt and GTDB r202, precision and F1 scores improved significantly with increasing CS, highlighting their robustness to stringent conditions. Recovery rates were mostly stable, indicating consistent detection of species under different CS settings. Crucially, the results show that a comprehensive reference database combined with a moderate CS (0.2 or 0.4) significantly improves classification accuracy and sensitivity. This finding underscores the need for careful selection of database and CS parameters tailored to specific scientific questions and available computational resources to optimize the results of metagenomic analyses.

An Optimized Miniaturized Filter-Aided Sample Preparation Method for Sensitive Cross-Linking Mass Spectrometry Analysis of Microscale Samples.

Cross-linking mass spectrometry (XL-MS) is a powerful tool for elucidating protein structures and protein-protein interactions (PPIs) at the global scale. However, sensitive XL-MS analysis of mass-limited samples remains challenging, due to serious sample loss during sample preparation of the low-abundance cross-linked peptides. Herein, an optimized miniaturized filter-aided sample preparation (O-MICROFASP) method was presented for sensitive XL-MS analysis of microscale samples. By systematically investigating and optimizing crucial experimental factors, this approach dramatically improves the XL identification of low and submicrogram samples. Compared with the conventional FASP method, more than 7.4 times cross-linked peptides were identified from single-shot analysis of 1 μg DSS cross-linked HeLa cell lysates (440 vs 59). The number of cross-linked peptides identified from 0.5 μg HeLa cell lysates was increased by 58% when further reducing the surface area of the filter to 0.058 mm2 in the microreactor. To deepen the identification coverage of XL-proteome, five different types of cross-linkers were used and each μg of cross-linked HeLa cell lysates was processed by O-MICROFASP integrated with tip-based strong cation exchange (SCX) fractionation. Up to 2741 unique cross-linked peptides were identified from the 5 μg HeLa cell lysates, representing 2579 unique K-K linkages on 1092 proteins. About 96% of intraprotein cross-links were within the maximal distance restraints of 26 Å, and 75% of the identified PPIs reported by the STRING database were with high confidence (scores ≥0.9), confirming the high validity of the identified cross-links for protein structural mapping and PPI analysis. This study demonstrates that O-MICROFASP is a universal and efficient method for proteome-wide XL-MS analysis of microscale samples with high sensitivity and reliability.

854 Bridging the Communication Gap in Surgery: A Quality Improvement Initiative to Improve Surgical Juniors’ Confidence in Navigating Difficult Conversations and Documentation

Abstract Aim In response to the challenges faced by surgical juniors in communicating with critically ill patients and their families, particularly after emergency surgery, we initiated a comprehensive two-pronged QIP. Our aims were 1) to enhance skills and confidence of surgical juniors in navigating difficult conversations, and 2) to improve documentation of post-operative patient and NOK discussions after ITU step-down to ward, in-line with NICE guidelines (CG83-1.13). Method A retrospective analysis of 14 laparotomy patients in September 2023, revealed that only 36% had documented NOK discussions 3-5 days post-operatively, falling short of best practice guidelines recommending documentation within 12 hours. Only 33% of the 10 patients requiring ITU admission, received documented NOK updates post-step-down to the ward. Interventions to address this included developing a communication skills workshop and implementing colored cards in notes folders to visually identify patients needing NOK discussions. Results Pre- and post-workshop surveys, utilizing a 1-to-5 scale, measured confidence levels of junior doctors for difficult discussions, with 13 and 9 respondents, respectively. High confidence scores (4 or 5) for different discussion types showed substantial improvement post-workshop: DNACPR (31% vs. 89%), end-of-life (31% vs. 55%), emotionally charged (31% vs. 78%). Furthermore, 100% of participants post-workshop were highly likely (4 or 5) to utilizing communication frameworks e.g., SPIES. Conclusions Our interventions have demonstrated a significant improvement in the confidence of surgical juniors when navigating challenging conversations. Our ongoing efforts include expanding our ‘communication in surgery’ workshop series, which we hope to present, and a re-audit of NOK discussion documentation to objectively measure positive changes post-intervention.

Smoking and vaping alter genes related to mechanisms of SARS-CoV-2 susceptibility and severity: a systematic review and meta-analysis.

Evidence for the impact of smoking on coronavirus disease 2019 (COVID-19) is contradictory, and there is little research on vaping. Here we provide greater clarity on mechanisms perturbed by tobacco cigarette, electronic cigarette and nicotine exposures that may impact the risks of infection and/or disease severity. Following Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines, the Ovid and Web of Science databases were searched. Study design and exposure-induced gene expression changes were extracted. Each study was quality assessed and higher confidence scores were assigned to genes consistently changed across multiple studies following the same exposure. These genes were used to explore pathways significantly altered following exposure. 125 studies provided data on 480 genes altered by exposure to tobacco cigarettes, e-cigarettes, nicotine or severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). Genes involved in both SARS-CoV-2 viral-entry and inflammation were changed following exposure. Pathway analysis revealed that many of those genes with high confidence scores are involved in common cellular processes relating to hyperinflammatory immune responses. Exposure to tobacco cigarettes, e-cigarettes or nicotine may therefore impact initial host-pathogen interactions and disease severity. Smokers and vapers of e-cigarettes with nicotine could potentially be at increased risk of SARS-CoV-2 infection, associated cytokine storm, and acute respiratory distress syndrome. However, further research is required, particularly on e-cigarettes, to determine the biological mechanisms involved in perturbation of viral-entry genes and host-pathogen interactions and subsequent responses within the respiratory tract. This will improve our physiological understanding of the impact of smoking and vaping on COVID-19, informing public health advice and providing improved guidance for management of SARS-CoV-2 and other respiratory viruses.

Development and validation of a self-attention network-based algorithm to detect mediastinal lesions on computed tomography images.

Diagnosis of mediastinal lesions on computed tomography (CT) images is challenging for radiologists, as numerous conditions can present as mass-like lesions at this site. This study aimed to develop a self-attention network-based algorithm to detect mediastinal lesions on CT images and to evaluate its efficacy in lesion detection. In this study, two separate large-scale open datasets [National Institutes of Health (NIH) DeepLesion and Medical Image Computing and Computer Assisted Intervention (MICCAI) 2022 Mediastinal Lesion Analysis (MELA) Challenge] were collected to develop a self-attention network-based algorithm for mediastinal lesion detection. We enrolled 921 abnormal CT images from the NIH DeepLesion dataset into the pretraining stage and 880 abnormal CT images from the MELA Challenge dataset into the model training and validation stages in a ratio of 8:2 at the patient level. The average precision (AP) and confidence score on lesion detection were evaluated in the validation set. Sensitivity to lesion detection was compared between the faster region-based convolutional neural network (R-CNN) model and the proposed model. The proposed model achieved an 89.3% AP score in mediastinal lesion detection and could identify comparably large lesions with a high confidence score >0.8. Moreover, the proposed model achieved a performance boost of almost 2% in the competition performance metric (CPM) compared to the faster R-CNN model. In addition, the proposed model can ensure an outstanding sensitivity with a relatively low false-positive rate by setting appropriate threshold values. The proposed model showed excellent performance in detecting mediastinal lesions on CT. Thus, it can drastically reduce radiologists' workload, improve their performance, and speed up the reporting time in everyday clinical practice.

Hypothesis: evidence that the PRS gene products of Saccharomyces cerevisiae support both PRPP synthesis and maintenance of cell wall integrity

The gene products of PRS1-PRS5 in Saccharomyces cerevisiae are responsible for the production of PRPP (5-phospho-D-ribosyl-α-1-pyrophosphate). However, it has been demonstrated that they are also involved in the cell wall integrity (CWI) signalling pathway as shown by protein–protein interactions (PPIs) with, for example Slt2, the MAP kinase of the CWI pathway. The following databases: SGD, BioGRID and Hit Predict, which collate PPIs from various research papers, have been scrutinized for evidence of PPIs between Prs1-Prs5 and components of the CWI pathway. The level of certainty in PPIs was verified by interaction scores available in the Hit Predict database revealing that well-documented interactions correspond with higher interaction scores and can be graded as high confidence interactions based on a score > 0.28, an annotation score ≥ 0.5 and a method-based high confidence score level of ≥ 0.485. Each of the Prs1-Prs5 polypeptides shows some degree of interaction with the CWI pathway. However, Prs5 has a vital role in the expression of FKS2 and Rlm1, previously only documented by reporter assay studies. This report emphasizes the importance of investigating interactions using more than one approach since every method has its limitations and the use of different methods, as described herein, provides complementary experimental and statistical data, thereby corroborating PPIs. Since the experimental data described so far are consistent with a link between PRPP synthetase and the CWI pathway, our aim was to demonstrate that these data are also supported by high-throughput bioinformatic analyses promoting our hypothesis that two of the five PRS-encoding genes contain information required for the maintenance of CWI by combining data from our targeted approach with relevant, unbiased data from high-throughput analyses.

Maximizing discrimination capability of knowledge distillation with energy function

To apply the latest computer vision techniques that require a large computational cost in real industrial applications, knowledge distillation methods (KDs) are essential. Existing logit-based KDs apply the constant temperature scaling to all samples in dataset, limiting the utilization of knowledge inherent in each sample individually. In our approach, we classify the dataset into two categories (i.e., low energy and high energy samples) based on their energy score. Through experiments, we have confirmed that low energy samples exhibit high confidence scores, indicating certain predictions, while high energy samples yield low confidence scores, meaning uncertain predictions. To distill optimal knowledge by adjusting non-target class predictions, we apply a higher temperature to low energy samples to create smoother distributions and a lower temperature to high energy samples to achieve sharper distributions. When compared to previous logit-based and feature-based methods, our energy-based KD (Energy KD) achieves better performance on various datasets. Especially, Energy KD shows significant improvements on CIFAR-100-LT and ImageNet datasets, which contain many challenging samples. Furthermore, we propose high energy-based data augmentation (HE-DA) for further improving the performance. We demonstrate that higher performance improvement could be achieved by augmenting only a portion of the dataset rather than the entire dataset, suggesting that it can be employed on resource-limited devices. To the best of our knowledge, this paper represents the first attempt to make use of energy function in knowledge distillation and data augmentation, and we believe it will greatly contribute to future research.

Deciphering the Systemic Impact of Herbal Medicines on Allergic Rhinitis: A Network Pharmacological Approach.

Allergic rhinitis (AR) is a systemic allergic disease that has a considerable impact on patients' quality of life. Current treatments include antihistamines and nasal steroids; however, their long-term use often causes undesirable side effects. In this context, traditional Asian medicine (TAM), with its multi-compound, multi-target herbal medicines (medicinal plants), offers a promising alternative. However, the complexity of these multi-compound traits poses challenges in understanding the overall mechanisms and efficacy of herbal medicines. Here, we demonstrate the efficacy and underlying mechanisms of these multi-compound herbal medicines specifically used for AR at a systemic level. We utilized a modified term frequency-inverse document frequency method to select AR-specific herbs and constructed an herb-compound-target network using reliable databases and computational methods, such as the Quantitative Estimate of Drug-likeness for compound filtering, STITCH database for compound-target interaction prediction (with a high confidence score threshold of 0.7), and DisGeNET and CTD databases for disease-gene association analysis. Through this network, we conducted AR-related targets and pathway analyses, as well as clustering analysis based on target-level information of the herbs. Gene ontology enrichment analysis was conducted using a protein-protein interaction network. Our research identified 14 AR-specific herbs and analyzed whether AR-specific herbs are highly related to previously known AR-related genes and pathways. AR-specific herbs were found to target several genes related to inflammation and AR pathogenesis, such as PTGS2, HRH1, and TBXA2R. Pathway analysis revealed that AR-specific herbs were associated with multiple AR-related pathways, including cytokine signaling, immune response, and allergic inflammation. Additionally, clustering analysis based on target similarity identified three distinct subgroups of AR-specific herbs, corroborated by a protein-protein interaction network. Group 1 herbs were associated with the regulation of inflammatory responses to antigenic stimuli, while Group 2 herbs were related to the detection of chemical stimuli involved in the sensory perception of bitter taste. Group 3 herbs were distinctly associated with antigen processing and presentation and NIK/NF-kappa B signaling. This study decodes the principles of TAM herbal configurations for AR using a network pharmacological approach, providing a holistic understanding of drug effects beyond specific pathways.

Structural insights into the molecular mechanism of phytoplasma immunodominant membrane protein.

Immunodominant membrane protein (IMP) is a prevalent membrane protein inphytoplasma and has been confirmed to be an F-actin-binding protein. However, the intricate molecular mechanisms that govern the function of IMP require further elucidation. In this study, the X-ray crystallographic structure ofIMP was determined and insights into its interaction with plant actin are provided. A comparative analysis with other proteins demonstrates that IMP shares structural homology with talin rod domain-containing protein 1 (TLNRD1), which also functions as an F-actin-binding protein. Subsequent molecular-docking studies of IMP and F-actin reveal that they possess complementary surfaces, suggesting a stable interaction. The low potential energy and high confidence score of the IMP-F-actin binding model indicate stable binding. Additionally, by employing immunoprecipitation and mass spectrometry, it was discovered that IMP serves as an interaction partner for the phytoplasmal effector causing phyllody 1 (PHYL1). It was then shown that both IMP and PHYL1 are highly expressed in the S2 stage of peanut witches' broom phytoplasma-infected Catharanthus roseus. The association between IMP and PHYL1 is substantiated through in vivo immunoprecipitation, an in vitro cross-linking assay and molecular-docking analysis. Collectively, these findings expand the current understanding of IMP interactions and enhance the comprehension of the interaction of IMP with plant F-actin. They also unveil a novel interaction pathway that may influence phytoplasma pathogenicity and host plant responses related to PHYL1. This discovery could pave the way for the development of new strategies to overcome phytoplasma-related plant diseases.

Low complex CNN for semiconductor wafer defect detection

Fabrication of semiconductor wafers is a complex process and chances of defect wafers are high. Because of defective wafers the circuit patterns will not be created correctly and it is necessary to identify them. Manual identification of defects are time consuming and expensive. Deep learning methods are widely used for defect detection. In this paper we propose a simple Convolutional Neural Network (CNN) model for classification of nine defects in wafers. A custom CNN consisting of 9 layers is used for the classification of defects as Center, Donut, Edge-Loc, Edge-Ring, Loc, Random, Scratch, Near-full, and None. Performance of the model is evaluated using WM-811K dataset. Results shows that the model classifies the defects with high confidence score and an accuracy of 99.1% is achieved using this method. Further, the convolution operation in the CNN is realized using Coordinate Rotation Digital Computer (CORDIC) algorithm. The model is implemented in Field Programmable Gate Arrays (FPGA) and proved less complex method and consume less computational power than conventional methods.

Perception, confidence, and willingness to respond to in-flight medical emergencies among medical students: a cross sectional study

Background In-flight medical emergencies (IMEs) are expected to increase as air travel normalized in the post-COVID-19 era. However, few studies have examined health professions students’ preparedness to respond to such emergencies. Therefore, this study aimed to investigate medical students’ knowledge, confidence, and willingness to assist during an IME in their internship program. Methods This cross-sectional survey utilized an online, self-administered questionnaire-based survey targeted at medical students at two medical colleges in Saudi Arabia. The questionnaire comprised three parts: sociodemographic characteristics, knowledge about aviation medicine (10 items), and confidence (7 items)/willingness (4 items) to assist during an IME. Odds Ratios (OR) and 95% Confidence Intervals (95%CI) were computed to detect potential associations between the knowledge levels and the other independent variables. Responses to confidence and willingness questions were scored on a 5-point Likert scale. Results Overall, 61.4% of participants had inadequate knowledge scores for providing care during an IME, and the proportion of participants did not differ between those who had or had not attended life support courses (60.4% vs. 66.7%, p > 0.99). Only frequency of air travel ≥ two times per year was associated with higher odds of adequate knowledge score [OR = 1.89 (95%CI 1.14–3.17), p = 0.02]. In addition, 93.3% of the participants had low, 6.3% had moderate, and 0.8% had high willingness scores, while 86.3% had low, 12.2% had moderate, and 1.5% had high confidence scores. There were no differences in the proportion of participants with low, moderate, and high willingness or confidence scores by attendance in life support courses. Conclusion Even though over 8 in 10 students in our study had previously attended life support courses, the overwhelming majority lacked the knowledge, confidence, and willingness to assist. Our study underscores the importance of teaching medical students about IMEs and their unique challenges before entering their 7th-year mandatory general internship.

Real-time object-removal tampering localization in surveillance videos by employing YOLO-V8.

Videos are considered as most trustworthy means of communication in the present digital era. The advancement in multimedia technology has made video content sharing and manipulation very easy. Hence, the video authenticity is a challenging task for the research community. Video forensics refer to uncovering the forgery traces. The detection of spatiotemporal object-removal forgery in surveillance videos is crucial for judicial forensics, as the presence of objects in the video has significant information as legal evidence. The author proposes a passive max-median averaging motion residual algorithm for revealing the forgery traces, successfully giving visible object-removal traces followed by a deep learning approach, YOLO-V8, for forged region localization. YOLO-V8 is the latest deep learning model, which has a wide scope for real-time application. The proposed method utilizes YOLO-V8 for object-removal forgery in surveillance videos. The network is trained on the SYSU-OBJFORG dataset for object-removal forged region localization in videos. The fine-tuned YOLO-V8 successfully classifies and localizes the object-removal tampered region with an F1-score of 0.99 and a precision of 0.99. The observed high confidence score of the bounding box around the forged region makes the model reliable. This fine-tuned YOLO-V8 would be a better choice in real-time applications as it solves the complex object-based forgery detection in videos. The performance of the proposed system is far better than the existing deep learning approach.

Comparison of CNN and SVM Methods on Web-based Skin Disease Classification Process

Skin, as the outermost layer of the body, is often in contact with bacteria, germs and viruses because of its most external position. According to statistics from the 2009 Indonesian Health Profile, skin illness is the third most common ailment seen in outpatient settings across the country's hospitals. Therefore, maintaining healthy skin is important because it protects the body's internal organs from injury and attack by pathogens. The development of image classification, such as the classification of skin diseases, has become a focus in the health sector. This research analyses the performance of Convolutional Neural Network (CNN) and Support Vector Machine (SVM) in web-based skin disease classification and overcomes the problem of imbalanced training data. With data augmentation and preprocess, this research improves data generalization and compares performance metrics such as Recall, Accuracy, and F1 Score. The results show that the average accuracy of CNN is 83.8%, while SVM reaches 81%. Although both models have high metrics for the normal class, other more complicated classes can only be handled by CNN with a value of more than 0.9. Apart from that, the CNN method also provides a higher Confidence Score than SVM, as well as a faster execution time. In conclusion, the CNN method is superior and recommended for skin disease classification based on web applications based on various performance test results.

Comparative Analysis of Deep Learning Models for Olive Detection on the Branch

The future of deep learning integration in agriculture holds great potential for advancing sustainable agricultural practices, precision agriculture and improved decision-making. With the rapid development of image processing and artificial intelligence technologies in recent years, deep learning has begun to play a major role in identifying agricultural pests and optimizing agricultural product marketing. However, there are challenges related to data quality, model scalability, and geographical limitations for widespread adoption of deep learning in agriculture. This study on Olive was conducted to improve the quality of the data set and to ensure more reliable training of object detection models. According to the result of the training process of YOLOv7 used in the study, it was concluded that it was characterized by decreasing loss values and showed an increase in the model's ability to detect objects correctly. It was observed that the other model, YOLOv8l, had a more effective learning capacity and a tendency to learn faster. The performance of both models was evaluated with various metrics, and it was determined that YOLOv8l had higher Precision, Recall, and mAP values. It was emphasized that YOLOv8l showed high performance even in low epoch numbers and can be preferred especially in cases where time and computational resources were limited. It was determined that YOLOv7 made detections in a wide confidence range, but had difficulty in detections with low confidence scores. It was observed that YOLOv8l made more stable and reliable detections with higher confidence scores. The metric data of the "YOLOv8l" model was found to be higher compared to other models. The F1 score of the YOLOv5l model was 92.337%, precision 96.568%, recall %88,462,mAP@0.5:0.65 value gave the highest score with 94.608%. This research on deep learning-based object detection models indicated that YOLOv8l showed superior performance compared to YOLOv7 and was a more reliable option for agricultural applications.

Mutual learning with reliable pseudo label for semi-supervised medical image segmentation

Semi-supervised learning has garnered significant interest as a method to alleviate the burden of data annotation. Recently, semi-supervised medical image segmentation has garnered significant interest that can alleviate the burden of densely annotated data. Substantial advancements have been achieved by integrating consistency-regularization and pseudo-labeling techniques. The quality of the pseudo-labels is crucial in this regard. Unreliable pseudo-labeling can result in the introduction of noise, leading the model to converge to suboptimal solutions. To address this issue, we propose learning from reliable pseudo-labels. In this paper, we tackle two critical questions in learning from reliable pseudo-labels: which pseudo-labels are reliable and how reliable are they? Specifically, we conduct a comparative analysis of two subnetworks to address both challenges. Initially, we compare the prediction confidence of the two subnetworks. A higher confidence score indicates a more reliable pseudo-label. Subsequently, we utilize intra-class similarity to assess the reliability of the pseudo-labels to address the second challenge. The greater the intra-class similarity of the predicted classes, the more reliable the pseudo-label. The subnetwork selectively incorporates knowledge imparted by the other subnetwork model, contingent on the reliability of the pseudo labels. By reducing the introduction of noise from unreliable pseudo-labels, we are able to improve the performance of segmentation. To demonstrate the superiority of our approach, we conducted an extensive set of experiments on three datasets: Left Atrium, Pancreas-CT and Brats-2019. The experimental results demonstrate that our approach achieves state-of-the-art performance. Code is available at: https://github.com/Jiawei0o0/mutual-learning-with-reliable-pseudo-labels.