Negative Samples Research Articles

A Method of Image Restoration for Distortion of Object in Water-Air Cross-Media

Uneven water-air media distribution or irregular liquid flow can cause changes in light propagation, leading to blurring and distortion of the extracted image, which presents a challenge for object recognition accuracy. To address these issues, this paper proposes a repair network to correct object image distortion in water-air cross-media. Firstly, convolutional combination performs feature extraction on water-air cross-media images, which retains the same features at the same scale and marks feature points with large differences. Then, an attention correction module for geometric lines is proposed to correct geometric lines in water-air cross-media images by comparing and sensing the marked feature points with large differences and utilizing the line similarity in positive and negative samples. Finally, the blurring artifact elimination module eliminates artifacts caused by image blurring and geometric line correction by using multiscale fusion of individual U-Net information streams. This completes the image restoration of object distortion under water-air cross-media. The proposed method is feasible and effective for restoring aberrated objects in water-air cross-media environments, with numerous experiments conducted on water-air cross-media image datasets.

Exploring the Performance of Continuous-Time Dynamic Link Prediction Algorithms

Dynamic Link Prediction (DLP) addresses the prediction of future links in evolving networks. However, accurately portraying the performance of DLP algorithms poses challenges that might impede progress in the field. Importantly, common evaluation pipelines usually calculate ranking or binary classification metrics, where the scores of observed interactions (positives) are compared with those of randomly generated ones (negatives). However, a single metric is not sufficient to fully capture the differences between DLP algorithms, and is prone to overly optimistic performance evaluation. Instead, an in-depth evaluation should reflect performance variations across different nodes, edges, and time segments. In this work, we contribute tools to perform such a comprehensive evaluation. (1) We propose Birth–Death diagrams, a simple but powerful visualization technique that illustrates the effect of time-based train–test splitting on the difficulty of DLP on a given dataset. (2) We describe an exhaustive taxonomy of negative sampling methods that can be used at evaluation time. (3) We carry out an empirical study of the effect of the different negative sampling strategies. Our comparison between heuristics and state-of-the-art memory-based methods on various real-world datasets confirms a strong effect of using different negative sampling strategies on the test area under the curve (AUC). Moreover, we conduct a visual exploration of the prediction, with additional insights on which different types of errors are prominent over time.

Rumor detection based on Attention Graph Adversarial Dual Contrast Learning.

It is becoming harder to tell rumors from non-rumors as social media becomes a key news source, which invites malicious manipulation that could do harm to the public's health or cause financial loss. When faced with situations when the session structure of comment sections is deliberately disrupted, traditional models do not handle them adequately. In order to do this, we provide a novel rumor detection architecture that combines dual comparison learning, adversarial training, and attention filters. We suggest the attention filter module to achieve the filtering of some dangerous comments as well as the filtering of some useless comments, allowing the nodes to enter the GAT graph neural network with greater structural information. The adversarial training module (ADV) simulates the occurrence of malicious comments through perturbation, giving the comments some defense against malicious comments. It also serves as a hard negative sample to aid double contrast learning (DCL), which aims to learn the differences between various comments, and incorporates the final loss in the form of a loss function to strengthen the model. According to experimental findings, our AGAD (Attention Graph Adversarial Dual Contrast Learning) model outperforms other cutting-edge algorithms on a number of rumor detection tasks. The code is available at https://github.com/icezhangGG/AGAD.git.

Multi-model combination in key steps for landslide susceptibility modeling and uncertainty analysis: a case study in Baoji City, China

Reliable landslide susceptibility maps are essential for geohazard risk management. However, the selection of models and methods in the landslide susceptibility modeling (LSM) process is subjective and can lead to uncertainties in susceptibility outcomes. This paper introduces a framework for LSM and uncertainty analysis that leverages multiple models in key steps. The framework dynamically operates and combines models at each LSM step to generate susceptibility evaluation results from various method combinations. The uncertainty of different model combinations is then analyzed by comparing these results. The framework’s effectiveness is validated using Baoji City as a case study, examining the impact of different attribute interval numbers (AIN), non-landslide negative sample selection methods, and prediction models on susceptibility prediction outcomes. The results of each group show that the highest AUC value is 0.963 (AIN = 12, buffer-controlled sampling, Random forest) is about 0.26 higher than the lowest (no AIN, low-slope controlled sampling, Support vector machine). The findings suggest that the combinations involving larger AIN values, buffer-controlled sampling, XGBoost, or Random forest model yield relatively high AUC accuracy and relatively low uncertainty in the susceptibility index.

Landslide susceptibility assessment based on frequency ratio and semi-supervised heterogeneous ensemble learning model.

Epistemic uncertainty in data-driven landslide susceptibility assessment often tends to be increased by the limited accuracy of an individual model, as well as uncertainties associated with the selection of non-landslide samples. To address these issues, this paper centers on the landslide disaster in Ji'an City, China, and proposes a heterogeneous ensemble learning method incorporating frequency ratio (FR) and semi-supervised sample expansion. Based on the superimposed results of 12 environmental factor frequency ratios (FFR), non-landslide samples were selected and input into light gradient boosting machine (LightGBM), random forest (RF), and convolutional neural network (CNN) models for prediction along with historical landslide samples. The predicted probability values are integrated by four heterogeneous ensemble strategies to expand samples from high-confidence results. The model's performance is evaluated using the area under the receiver operating characteristic curve (AUC), partition frequency ratio (PFR), and other verification methods. The results demonstrate that the negative sample based on FFR sampling is more accurate than the random sampling method, and the FR-SSELR model based on frequency ratio sampling and semi-supervised ensemble strategy exhibits the highest performance (AUC = 0.971, ACC = 0.941). A more reasonable landslide susceptibility map was drawn based on this model, with the lowest percentage of landslides in the low and very low susceptibility zones (sum of PFR = 0.194), as well as the highest percentage of landslides in the high and very high susceptibility zones (sum of PFR = 6.800). Furthermore, the FR-SSELR model improved economic benefits by 3.82-14.2%, offering valuable guidance for decision-making regarding landslide management and the sustainability of Ji'an City.

Prediction of blood–brain barrier penetrating peptides based on data augmentation with Augur

BackgroundThe blood–brain barrier serves as a critical interface between the bloodstream and brain tissue, mainly composed of pericytes, neurons, endothelial cells, and tightly connected basal membranes. It plays a pivotal role in safeguarding brain from harmful substances, thus protecting the integrity of the nervous system and preserving overall brain homeostasis. However, this remarkable selective transmission also poses a formidable challenge in the realm of central nervous system diseases treatment, hindering the delivery of large-molecule drugs into the brain. In response to this challenge, many researchers have devoted themselves to developing drug delivery systems capable of breaching the blood–brain barrier. Among these, blood–brain barrier penetrating peptides have emerged as promising candidates. These peptides had the advantages of high biosafety, ease of synthesis, and exceptional penetration efficiency, making them an effective drug delivery solution. While previous studies have developed a few prediction models for blood–brain barrier penetrating peptides, their performance has often been hampered by issue of limited positive data.ResultsIn this study, we present Augur, a novel prediction model using borderline-SMOTE-based data augmentation and machine learning. we extract highly interpretable physicochemical properties of blood–brain barrier penetrating peptides while solving the issues of small sample size and imbalance of positive and negative samples. Experimental results demonstrate the superior prediction performance of Augur with an AUC value of 0.932 on the training set and 0.931 on the independent test set.ConclusionsThis newly developed Augur model demonstrates superior performance in predicting blood–brain barrier penetrating peptides, offering valuable insights for drug development targeting neurological disorders. This breakthrough may enhance the efficiency of peptide-based drug discovery and pave the way for innovative treatment strategies for central nervous system diseases.

GAN-based self-supervised message passing graph representation learning

Graph representation learning aims to learn embedding representations of nodes in graph-structured data through supervised or unsupervised methods. However, for unlabeled or sparsely labeled graph datasets, conventional supervised approaches fail to perform adequately, and existing unsupervised methods have difficulty achieving high accuracy on multiple downstream tasks such as node classification and link prediction at lower embedding dimensions. To address this issue, we propose a novel self-supervised graph representation learning approach called GAN-based Message Passing Graph Representation Learning. This approach significantly improves the accuracy of node embeddings without relying on label information. Specifically, GMP-GL has developed a generator based on neighbor-weighted random walk to obtain positive and negative node sample pairs, as well as a discriminator that distinguishes whether a real connection exists between node pairs. Through adversarial learning, GMP-GL automatically learns node embeddings. Furthermore, GMP-GL optimizes the training outcomes of the generative adversarial model through self-supervised embedding aggregation learning, thus introducing rich structural information from the graph for the subsequent rounds of generative adversarial training. In addition, GMP-GL employs multiple autoencoders to generate and filter original embeddings, ensuring that the pre-embeddings preserve a significant amount of original feature information for formal learning. The study selects six real-world datasets and conducts comparative experiments with ten SOTA baseline methods. The experimental results demonstrate that the proposed model not only surpasses previous unsupervised methods but also matches or even exceeds the performance of supervised methods. This indicates the great potential of the proposed approach for representation learning tasks on unlabeled graph datasets and its effective applicability to real-world problems.

Polymeric optical fiber biosensor with PAMAM dendrimer-based surface modification and PlGF detection for pre-eclampsia diagnosis

Pre-eclampsia (PE) is a life-threatening complication that occurs during pregnancy, affecting a large number of pregnant women and newborns worldwide. Rapid, on-site and affordable screening of PE at an early stage is necessary to ensure timely treatment and minimize both maternal and neonatal morbidity and mortality rates. Placental growth factor (PlGF) is an angiogenic blood biomarker used for PE diagnosis. Herein, we report the plasmonic fiber optic absorbance biosensor (P-FAB) strategy for detecting PlGF at femtomolar concentration using polymethyl methacrylate (PMMA) based U-bent polymeric optical fiber (POF) sensor probes. A novel poly(amidoamine) (PAMAM) dendrimer based PMMA surface modification is established to obtain a greater immobilization of the bioreceptors compared to a linear molecule like hexamethylenediamine (HMDA). Plasmonic sandwich immunoassay was realized by immobilizing the mouse anti-PlGF (3H1) on the U-bent POF sensor probe surface and gold nanoparticles (AuNP) labels conjugated with mouse anti-PlGF (6H9). The POF sensor probes could measure PlGF within 30 min using the P-FAB strategy. The limit-of-detection (LoD) was found to be 0.19 pg/mL and 0.57 pg/mL in phosphate-buffered saline and 10× diluted serum, respectively. The clinical sample testing, with eleven positive and eleven negative preeclamptic pregnancy samples, successfully confirmed the accuracy, reliability, specificity, and sensitivity of the P-FAB based POF sensor platform, thereby paving the way for cost-effective technology for PlGF detection and its potential for pre-eclampsia diagnosis.

Low-quality image object detection based on reinforcement learning adaptive enhancement

The task of object detection for low-quality images in multiple scenarios remains a challenging task. To alleviate this problem, a novel approach to improve the detection performance of the model on low-quality images using deep reinforcement learning is proposed. The approach differs from methods that involve data labeling and retraining, as the detection results of the model are utilized to formulate a reward function. This enables deep reinforcement learning to transform low-quality images into high-quality images through image enhancement, ultimately leading to an improvement in the model's recognition performance on low-quality images. An image enhancement tool chain (IETC) has been prepared, offering a flexible and lightweight algorithm selection for the Dueling Deep Q-Network-based Tool Selector (DDQN-TS). By adding a “Pass” option, Enable DDQN-TS to learn the relative relationship between performing and not performing image enhancement. To mitigate the potential inconsistency between the output of the detection model and the actual situation caused by negative samples, which can lead to incorrect decisions made by DDQN-TS, the introduction of a “thresh” parameter regulates the level of intervention in target detection by DIE. The experimental results demonstrate the effectiveness and portability of the DIE model in environments with noise, fog, and uneven brightness.

Advancing human genotyping: The Infinium HTS iSelect Custom microarray panel (Rita) development study

Single Nucleotide Polymorphisms (SNPs), as the most prevalent type of variation in the human genome, play a pivotal role in influencing human traits. They are extensively utilized in diverse fields such as population genetics, forensic science, and genetic medicine. This study focuses on the 'Rita' BeadChip, a custom SNP microarray panel developed using Illumina Infinium HTS technology. Designed for high-throughput genotyping, the panel facilitates the analysis of over 4000 markers efficiently and cost-effectively. After careful clustering performed on a set of 1000 samples, an evaluation of the Rita panel was undertaken, assessing its sensitivity, repeatability, reproducibility, precision, accuracy, and resistance to contamination. The panel’s performance was evaluated in various scenarios, including sex estimation and parental relationship assessment, using GenomeStudio data analysis software. Findings show that over 95 % of the custom BeadChip assay markers were successful, with better performance of transitions over other mutations, and a considerably lower success rate for Y chromosome loci. An exceptional call rate exceeding 99 % was demonstrated for control samples, even with DNA input as low as 0.781 ng. Call rates above 80 % were still obtained with DNA quantities under 0.1 ng, indicating high sensitivity and suitability for forensic applications where DNA quantity is often limited. Repeatability, reproducibility, and precision studies revealed consistency of the panel's performance across different batches and operators, with no significant deviations in call rates or genotyping results. Accuracy assessments, involving comparison with multiple available genetic databases, including the 1000 Genome Project and HapMap, denoted over 99 % concordance, establishing the Rita panel's reliability in genotyping. The contamination study revealed insights into background noise and allowed the definition of thresholds for sample quality evaluation. Multiple metrics for differentiating between negative controls and true samples were highlighted, increasing the reliability of the obtained results. The sex estimation tool in GenomeStudio proved highly effective, correctly assigning sex in all samples with autosomal loci call rates above 97 %. The parental relationship assessment of family trios highlighted the utility of GenomeStudio in identifying genotyping errors or potential Mendelian inconsistencies, promoting the application of arrays such as Rita in kinship testing. Overall, this evaluation confirms the Rita microarray as a robust, high-throughput genotyping tool, underscoring its potential in genetic research and forensic applications. With its custom content and adaptable design, it not only meets current genotyping demands but also opens avenues for further research and application expansion in the field of genetic analysis.

An effective object detector via diffused graphic large selective kernel with one-to-few labelling strategy for small-scaled crop diseases detection

Amid the swift evolution of advanced technologies such as autonomous robotics, edge computing, and the Internet of Things, Intelligent Agriculture Management Systems (IAMS) have become increasingly pivotal in contemporary smart agriculture. One vital task, crop disease monitoring, significantly contributes to ensuring agricultural production quality and food supply security. However, IAMS encounters a particular challenge: due to the limited computational capacity of agricultural equipment and the scarcity of available training datasets, general detection models struggle to achieve anticipated performance in agricultural environments, especially when dealing with small-scaled targets. In this paper, we propose a diffusion learning detector named CropDetdiff, designed for disease detection in agricultural environments where training samples are limited. We have developed a backbone network based on dynamic convolutional kernels, which also embeds a graph attention network in its final stage, enhancing feature extraction capabilities for small targets. Additionally, we employ a fuzzy label assignment strategy to balance positive and negative samples and devise a new loss function, HIoU, to supervise the detection of small targets. A series of experiments on two crop disease datasets validate that CropDetDiff achieves an optimal balance between detection performance and operational efficiency. Specifically, the proposed model has achieved a mean Average Precision (mAP) of 73.2% and 40.1% on the two datasets, improving by at least 3.7% and 2.9% over other detectors, respectively, and there is a significant improvement in small targets particularly, thus achieving optimal performance. Additionally, the proposed model exhibits commendable performance in terms of parameter quantity and execution velocity. These results substantiate the efficacy and stability of CropDetDiff, underscoring its suitability for real-world applications within IAMS.

Supervised contrastive learning for graph representation enhancement

Graph Neural Networks (GNNs) have exhibited significant success in various applications, but they face challenges when labeled nodes are limited. A novel self-supervised learning paradigm has emerged, enabling GNN training without labeled nodes and even surpassing GNNs with limited labeled data. However, self-supervised methods lack class-discriminative node representations due to the absence of labeled information during training. In this paper, we exploit a supervised graph contrastive learning approach (SGCL) framework to tackle the issue of limited labeled nodes, ensuring coherent grouping of nodes within the same class. We propose augmentation techniques based on a novel centrality function to highlight important topological structures. Additionally, we introduce a supervised contrastive learning method that removes the necessity for negative samples and simplifies complex elements effortlessly. Our approach combines supervised contrastive loss and node similarity regularization while achieving consistent grouping of unlabeled nodes with labeled ones. Furthermore, we utilize the pseudo-labeling technique to propagate label information to distant nodes and address the underfitting problem, especially with low-degree nodes. Experimental results on real-world graphs demonstrate that SGCL outperforms both semi-supervised and self-supervised methods in node classification.

Structural self-contrast learning based on adaptive weighted negative samples for facial expression recognition

Structural self-contrast learning based on adaptive weighted negative samples for facial expression recognition

Detection of intrathecal IgG antibody for varicella and measles diagnosis by evaluation and comparison of a commercial IgG chemiluminescent immunoassay with two ELISAs.

Analyse alternative methods of intrathecal antibody detection by comparing chemiluminescent immunoassay (CLIA) and enzyme-linked immunosorbent assay (ELISA) techniques to determine if CLIA can replace ELISA in the diagnosis of CNS infections. A panel of 280 paired samples-cerebrospinal fluid (CSF) and serum-with known antibody reactivities (Varicella, n = 60; Measles, n = 120) and negative samples (n = 100) were used to evaluate the performance of six serological test kits (Enzygnost, VirClia®, and Serion ELISA (Measles and Variella). For Measles virus IgG, the VirClia® IgG monotest revealed 97% and 94% positive and negative agreement to the Enzygnost as reference test,respectively. In contrast, Serion ELISA kits yielded values of 18% and 90%. For theVaricella Zoster virus (VZV) IgG, the VirClia® IgG monotest showed 97% and 90% positive and negative agreement compared to Enzygnost. The Serion ELISA kits showed values of 55% and 86%, respectively. ROC analysis revealed that the areas under the curve for Measles and VZV IgGs were 0.7 and 0.852, respectively, using the Serion kit, and 0.963 and 0.955, for Vircell S.L CLIA technique. VirClia® monotest values were calculated using an antibody index cut-off of 1.3. The findings indicate that CLIA testing can improve antibody detection in CSF samples, aiding the diagnosis of infectious neurological impairments.

Algorithm of face anti-spoofing based on pseudo-negative features generation.

Despite advancements in face anti-spoofing technology, attackers continue to pose challenges with their evolving deceptive methods. This is primarily due to the increased complexity of their attacks, coupled with a diversity in presentation modes, acquisition devices, and prosthetic materials. Furthermore, the scarcity of negative sample data exacerbates the situation by causing domain shift issues and impeding robust generalization. Hence, there is a pressing need for more effective cross-domain approaches to bolster the model's capability to generalize across different scenarios. This method improves the effectiveness of face anti-spoofing systems by analyzing pseudo-negative sample features, expanding the training dataset, and boosting cross-domain generalization. By generating pseudo-negative features with a new algorithm and aligning these features with the use of KL divergence loss, we enrich the negative sample dataset, aiding the training of a more robust feature classifier and broadening the range of attacks that the system can defend against. Through experiments on four public datasets (MSU-MFSD, OULU-NPU, Replay-Attack, and CASIA-FASD), we assess the model's performance within and across datasets by controlling variables. Our method delivers positive results in multiple experiments, including those conducted on smaller datasets. Through controlled experiments, we demonstrate the effectiveness of our method. Furthermore, our approach consistently yields favorable results in both intra-dataset and cross-dataset evaluations, thereby highlighting its excellent generalization capabilities. The superior performance on small datasets further underscores our method's remarkable ability to handle unseen data beyond the training set.

Reinvestigating the performance of artificial intelligence classification algorithms on COVID-19 X-Ray and CT images

There are concerns that artificial intelligence (AI) algorithms may create underdiagnosis bias by mislabeling patient individuals with certain attributes (e.g., female and young) as healthy. Addressing this bias is crucial given the urgent need for AI diagnostics facing rapidly spreading infectious diseases like COVID-19. We find the prevalent AI diagnostic models show an underdiagnosis rate among specific patient populations, and the underdiagnosis rate is higher in some intersectional specific patient populations (for example, females aged 20-40 years). Additionally, we find training AI models on heterogeneous datasets (positive and negative samples from different datasets) may lead to poor model generalization. The model's classification performance varies significantly across test sets, with the accuracy of the better performance being over 40% higher than that of the poor performance. In conclusion, we developed an AI bias analysis pipeline to help researchers recognize and address biases that impact medical equality and ethics.

Ultrafast Molecular Diagnosis‐Based Solid‐Phase Photonic PCR for Respiratory Pathogen Variant Discrimination

AbstractThe need for an accurate and rapid diagnosis platform for highly contagious viruses, such as pandemic viruses, has rapidly increased. Conventional diagnostic methods in the initial stage of quarantine for the prevention of epidemic spread have encountered limitations such as a long operation time, occupation of a large space, and high cost per single operation. Here, a highly sensitive and selective portable diagnostic platform based on solid‐phase photonic polymerase chain reaction (SP‐PCR) is reported. The thermal cycle is operated by the photothermal energy converted from the photon energy of a light‐emitting diode (LED) through a gold layer to phonons. The formation of an amplicon, which is immobilized through N‐heterocyclic carbene, is confirmed through fluorescence on the surface of the miniaturized SP‐PCR chip. The practical applicability of the newly developed diagnosis platform for the recognition of the presence of a SARS‐CoV‐2‐specific genome sequence is demonstrated utilizing 100 clinical samples, including 72 COVID‐19‐positive clinical samples and 28 negative samples.

Improved YOLOv5 Based on Multi-Strategy Integration for Multi-Category Wind Turbine Surface Defect Detection

Wind energy is a renewable resource with abundant reserves, and its sustainable development and utilization are crucial. The components of wind turbines, particularly the blades and various surfaces, require meticulous defect detection and maintenance due to their significance. The operational status of wind turbine generators directly impacts the efficiency and safe operation of wind farms. Traditional surface defect detection methods for wind turbines often involve manual operations, which suffer from issues such as high subjectivity, elevated risks, low accuracy, and inefficiency. The emergence of computer vision technologies based on deep learning has provided a novel approach to surface defect detection in wind turbines. However, existing datasets designed for wind turbine surface defects exhibit overall category scarcity and an imbalance in samples between categories. The algorithms designed face challenges, with low detection rates for small samples. Hence, this study first constructs a benchmark dataset for wind turbine surface defects comprising seven categories that encompass all common surface defects. Simultaneously, a wind turbine surface defect detection algorithm based on improved YOLOv5 is designed. Initially, a multi-scale copy-paste data augmentation method is proposed, introducing scale factors to randomly resize the bounding boxes before copy-pasting. This alleviates sample imbalances and significantly enhances the algorithm’s detection capabilities for targets of different sizes. Subsequently, a dynamic label assignment strategy based on the Hungarian algorithm is introduced that calculates the matching costs by weighing different losses, enhancing the network’s ability to learn positive and negative samples. To address overfitting and misrecognition resulting from strong data augmentation, a two-stage progressive training method is proposed, aiding the model’s natural convergence and improving generalization performance. Furthermore, a multi-scenario negative-sample-guided learning method is introduced that involves incorporating unlabeled background images from various scenarios into training, guiding the model to learn negative samples and reducing misrecognition. Finally, slicing-aided hyper inference is introduced, facilitating large-scale inference for wind turbine surface defects in actual industrial scenarios. The improved algorithm demonstrates a 3.1% increase in the mean average precision (mAP) on the custom dataset, achieving 95.7% accuracy in mAP_50 (the IoU threshold is half of the mAP). Notably, the mAPs for small, medium, and large targets increase by 18.6%, 16.4%, and 6.8%, respectively. The experimental results indicate that the enhanced algorithm exhibits high detection accuracy, providing a new and more efficient solution for the field of wind turbine surface defect detection.

Performance Assessment of Treponemal and Nontreponemal Tests for the Diagnosis of Acquired Syphilis.

There are a variety of nontreponemal test (NTT) and treponemal test (TT) kits for the serologic diagnosis of syphilis. Because of the complexity of the infection (multiple clinical stages) and the different antigens used in these kits, a systematic evaluation of the accuracy of the currently available commercial tests is warranted. Our objective was to evaluate the performance of commercially available tests for the diagnosis of syphilis infection. In this study, we analyzed one NTT (Venereal Disease Research Laboratory [VDRL] test, Wiener Laboratories, Rosario, Argentina) and two TTs (fluorescent treponemal antibody absorption [FTA-ABS] test, Euroimmun, Lübeck, Germany, and syphilis recombinant ELISA v. 4.0 test [ELISA], Wiener Laboratories, Rosario, Argentina) using a panel of 187 samples, including serum samples from 31 individuals with primary syphilis, 77 with secondary syphilis, and 79 with latent syphilis. An additional 192 samples from uninfected individuals and 323 serum samples from individuals with other diseases were included. The sensitivities of the VDRL, ELISA, and FTA-ABS tests were 97.9%, 100%, and 96.3%, respectively. The VDRL and ELISA tests showed a specificity of 100%, and the FTA-ABS test showed a specificity of 99.5%. Accuracy was 98.9% for the VDRL test, 100% for the ELISA, and 97.9% for the FTA-ABS test. For primary, secondary, and latent syphilis, the ELISA achieved a diagnostic performance of 100%, whereas the sensitivity for the VDRL and FTA-ABS tests ranged from 96.8% to 98.7% and 93.7% to 98.7%, respectively. No difference was observed when the tests were used as traditional or reverse algorithms. In general, all three tests are able to discriminate positive and negative samples for syphilis, regardless of the diagnostic algorithm.

ESF-YOLO: an accurate and universal object detector based on neural networks.

As an excellent single-stage object detector based on neural networks, YOLOv5 has found extensive applications in the industrial domain; however, it still exhibits certain design limitations. To address these issues, this paper proposes Efficient Scale Fusion YOLO (ESF-YOLO). Firstly, the Multi-Sampling Conv Module (MSCM) is designed, which enhances the backbone network's learning capability for low-level features through multi-scale receptive fields and cross-scale feature fusion. Secondly, to tackle occlusion issues, a new Block-wise Channel Attention Module (BCAM) is designed, assigning greater weights to channels corresponding to critical information. Next, a lightweight Decoupled Head (LD-Head) is devised. Additionally, the loss function is redesigned to address asynchrony between labels and confidences, alleviating the imbalance between positive and negative samples during the neural network training. Finally, an adaptive scale factor for Intersection over Union (IoU) calculation is innovatively proposed, adjusting bounding box sizes adaptively to accommodate targets of different sizes in the dataset. Experimental results on the SODA10M and CBIA8K datasets demonstrate that ESF-YOLO increases Average Precision at 0.50 IoU (AP50) by 3.93 and 2.24%, Average Precision at 0.75 IoU (AP75) by 4.77 and 4.85%, and mean Average Precision (mAP) by 4 and 5.39%, respectively, validating the model's broad applicability.