Enhance Detection Accuracy Research Articles

Multimodal Fake News Detection with Contrastive Learning and Optimal Transport

IntroductionThe proliferation of social media platforms has facilitated the spread of fake news, posing significant risks to public perception and societal stability. Existing methods for multimodal fake news detection have made important progress in combining textual and visual information but still face challenges in effectively aligning and merging these different types of data. These challenges often result in incomplete or inaccurate feature representations, thereby limiting overall performance.MethodsTo address these limitations, we propose a novel framework named MCOT (Multimodal Fake News Detection with Contrastive Learning and Optimal Transport). MCOT integrates textual and visual information through three key components: cross-modal attention mechanism, contrastive learning, and optimal transport. Specifically, we first use cross-modal attention mechanism to enhance the interaction between text and image features. Then, we employ contrastive learning to align related embeddings while distinguishing unrelated pairs, and we apply optimal transport to refine the alignment of feature distributions across modalities.ResultsThis integrated approach results in more precise and robust feature representations, thus enhancing detection accuracy. Experimental results on two public datasets demonstrate that the proposed MCOT outperforms state-of-the-art methods.DiscussionOur future work will focus on improving its generalization and expanding its capabilities to additional modalities.

A highly efficient tunnel lining crack detection model based on Mini-Unet.

The accurate detection of tunnel lining cracks and prompt identification of their primary causes are critical for maintaining tunnel availability. The advancement of deep learning, particularly in the domain of convolutional neural network (CNN) for image segmentation, has made tunnel lining crack detection more feasible. However, the CNN-based technique for tunnel lining crack detection commonly prioritizes increasing algorithmic complexity to enhance detection accuracy, posing a challenge in balancing the accuracy of detection and the efficiency of the algorithm. Motivated by the superior performance of Unet in image segmentation, this paper proposes a lightweight tunnel lining crack detection model named Mini-Unet, which refined the Unet architecture and utilized depthwise separable convolutions (DSConv) to replace some standard convolution layers. In the optimization of the proposed model parameters, applying a hybrid loss function that integrated dice loss and cross-entropy loss effectively tackled the imbalance between crack and background categories. Several models were set up to contrast with Mini-Unet and the experimental results were analyzed. Mini-Unet achieves a mean intersection over union (MIoU) of 60.76%, a mean precision of 84.18%, and a frame per second (FPS) of 5.635, respectively. Mini-Unet outperforms several mainstream models, enabling rapid detection while maintaining identified accuracy and facilitating the practical application of AI power for real-time tunnel lining crack detection.

LIDeepDet: Deepfake Detection via Image Decomposition and Advanced Lighting Information Analysis

The proliferation of AI-generated content (AIGC) has empowered non-experts to create highly realistic Deepfake images and videos using user-friendly software, posing significant challenges to the legal system, particularly in criminal investigations, court proceedings, and accident analyses. The absence of reliable Deepfake verification methods threatens the integrity of legal processes. In response, researchers have explored deep forgery detection, proposing various forensic techniques. However, the swift evolution of deep forgery creation and the limited generalizability of current detection methods impede practical application. We introduce a new deep forgery detection method that utilizes image decomposition and lighting inconsistency. By exploiting inherent discrepancies in imaging environments between genuine and fabricated images, this method extracts robust lighting cues and mitigates disturbances from environmental factors, revealing deeper-level alterations. A crucial element is the lighting information feature extractor, designed according to color constancy principles, to identify inconsistencies in lighting conditions. To address lighting variations, we employ a face material feature extractor using Pattern of Local Gravitational Force (PLGF), which selectively processes image patterns with defined convolutional masks to isolate and focus on reflectance coefficients, rich in textural details essential for forgery detection. Utilizing the Lambertian lighting model, we generate lighting direction vectors across frames to provide temporal context for detection. This framework processes RGB images, face reflectance maps, lighting features, and lighting direction vectors as multi-channel inputs, applying a cross-attention mechanism at the feature level to enhance detection accuracy and adaptability. Experimental results show that our proposed method performs exceptionally well and is widely applicable across multiple datasets, underscoring its importance in advancing deep forgery detection.

Determining the optimal time window to detect emerald ash borer damage for effective management.

The emerald ash borer (EAB) is an invasive pest of global concern. Accurate detection of EAB is crucial for effective management. Traditional field surveys fail to meet large-scale monitoring requirements. Remote sensing methods offer a potential solution, but the phenological decline of ash trees may obscure the remote sensing features for detecting EAB. Therefore, determining the timing of leaf abscission caused by EAB before phenology is crucial for effective detection. We collected time-series data of Leaf Area Index (LAI), leaf sizes, and hyperspectral images of damaged ash trees throughout the growing season to determine the optimal detecting time window for EAB detection using field surveys or remote sensing techniques. Significant differences in LAI and leaf size were observed throughout the growing season among ash trees with different EAB infestation degrees, providing a basis for small-scale field surveys. However, in May and June, the hyperspectral reflectance showed no variation. The difference began to appear in July and became apparent from August to October. By October, severely EAB-infested ash trees had almost completely defoliated. Machine learning classification results showed that accuracies after July were higher than before July. After July, the highest classification accuracy reached 100%, while the highest accuracy before July was only 88.57%. Selecting the optimal monitoring time significantly enhanced detection accuracy. The optimal period for field surveys is from May to November, whereas for remote sensing it is from August to October. Identifying the optimal months enables us to achieve more efficient decision-making and management. © 2024 Society of Chemical Industry.

Hierarchical Image Quality Improvement Based on Illumination, Resolution, and Noise Factors for Improving Object Detection

Object detection performance is significantly impacted by image quality factors such as illumination, resolution, and noise. This paper proposes a hierarchical image quality improvement process that dynamically prioritizes these factors based on severity, enhancing detection accuracy in diverse conditions. The process evaluates each factor—illumination, resolution, and noise—using discriminators that analyze brightness, edge strength, and noise levels. Improvements are applied iteratively with an adaptive weight update mechanism that adjusts factor importance based on improvement effectiveness. Following each improvement, a quality assessment is conducted, updating weights to fine-tune subsequent adjustments. This allows the process to learn optimal parameters for varying conditions, enhancing adaptability. The image improved through the proposed process shows improved quality through quality index (PSNR, SSIM) evaluation, and the object detection accuracy is significantly improved when the performance is measured using deep learning models called YOLOv8 and RT-DETR. The detection rate is improved by 7% for the ‘Bottle’ object in a high-light environment, and by 4% and 2.5% for the ‘Bicycle’ and ‘Car’ objects in a low-light environment, respectively. Additionally, segmentation accuracy saw a 9.45% gain, supporting the effectiveness of this method in real-world applications.

Detection and identification of centipedes based on deep learning.

The quantification of centipede populations is one of the key measures in achieving intelligent management of edible centipedes and promoting the upgrade of the rural centipede industry chain. However, current centipede counting techniques still face several challenges, including low detection accuracy, large model size, and difficulty in deployment on mobile devices. These challenges have limited existing network models to the experimental stage, preventing their practical application. To tackle the identified challenges, this study introduces a lightweight centipede detection model (FCM-YOLO), which enhances detection performance while ensuring fast processing and broad applicability. Based on the YOLOv5s framework, this model incorporates the C3FS module, resulting in fewer parameters and increased detection speed. Additionally, it integrates an attention module (CBAM) to suppress irrelevant information and improve target focus, thus enhancing detection accuracy. Furthermore, to enhance the precision of bounding box positioning, this study proposes a new loss function, CMPDIOU, for bounding box loss. Experimental results show that FCM-YOLO, while reducing parameter size, achieves an improved detection accuracy of 97.4% (2.7% higher than YOLOv5s) and reduces floating-point operations (FLOPs) to 11.5G (4.3G lower than YOLOv5s). In summary, this paper provides novel insights into the detection and enumeration of centipedes, contributing to the advancement of intelligent agricultural practices.

A new target color adaptive graying and segmentation method for gear contact spot detection

Abstract Gear contact spot plays a crucial role in evaluating gear mesh quality. Traditionally, tooth surfaces of gear pairs have been manually brushed with red lead powder and visually inspected after a certain operating period. To enhance detection accuracy and efficiency, computer vision has emerged as an appealing approach for gear contact spot detection. However, determining the graying weights and segmentation threshold poses challenges, particularly in non-ideal illumination environments. To address this problem, this paper proposes a new target color adaptive graying and segmentation (TC-AGS) method for the gear contact spot detection. With the proposed method, the illumination scene is firstly distinguished by the blue color component weight due to its high sensitivity to illumination intensity. Then, an improved adaptively graying algorithm for target color is employed to determine the graying weights for RGB, ensuring the maximum proportion of the desired red color. Finally, either a global threshold method or a local threshold method is selected for image segmentation and contact spot detection according to the illumination intensity. To validate the effectiveness of the proposed method, comprehensive simulation and experimental tests were conducted. The results demonstrate that the proposed method achieves better gear contact spot detection performance than traditional methods. It exhibits higher extraction accuracy under different illumination environments and even in the presence of blurred image sources.

Lead acetate-based test strip method for rapid and quantitative detection of residual sulfur dioxide in Chinese herbal medicines.

To prevent the residual sulfur dioxide in Chinese herbal medicines (CHM) caused by sulfur fumigation, which may lead to severe health issues, there is an urgent need for a rapid and quantitative detection technique. Sodium borohydride was used as a reducing agent to convert sulfur dioxide into hydrogen sulfide, which was then detected using lead acetate test strip. An accurate testing apparatus was designed, consisting of reaction bottle cap, reaction bottle, lead acetate test strip, and sulfur dioxide detector. The effect of different reaction variables on detection, including reductant quality, pH of initial media, reaction time, lead acetate concentration, and membrane type was investigated. The optimal conditions were determined by orthogonal experiments. The reaction membrane type and lead acetate concentration on the membrane were optimized to enhance detection accuracy. Standardized gray cards were fabricated and used to calibrate the detector. The detection system demonstrated an exceptional linear correlation (r2 = 0.9992), with a linear detection range of 0-750 mg·kg-1. The colored substances and sulfur-containing substances within the matrix of CHM did not affect the detection results. Therefore, the detection method exhibited superior accuracy and stability. The proposed technique proved to be swift, reliable, and provides a straightforward and convenient approach for the quantitative determination of sulfur dioxide in CHMs. The results of this work may provide insights into the development of test strips for quantitative detection.

Impact of RAW vs. RGB images on low-light object detection in autonomous driving

Abstract. In the field of autonomous driving, object detection in low-light conditions presents a critical challenge. Traditional detection algorithms, reliant on RGB imagery, often suffer from diminished performance under inadequate illumination. In contrast, RAW images, which retain a higher level of intrinsic image information, have the potential to enhance detection accuracy. Therefore, the paper aims to assess whether models trained on RAW images exhibit enhanced object detection performance in low-light scenarios. To this end, experiments are conducted utilizing the Low-light Object Detection (LOD) dataset, which includes paired RAW and RGB images, with variations in lighting simulated through different exposure times. Two distinct datasets are constructed: one consisting of RAW-normal and RAW-dark images, and another comprising RGB-normal and RGB-dark images. The models evaluated in this research include YOLOv8, Faster R-CNN (Region-based Convolutional Neural Network), and EfficientDet, selected to ensure robustness and generalizability of the findings. The results demonstrate that models trained on RAW images significantly outperform those trained on RGB images, thus exhibiting higher accuracy both in general and under low-light conditions. The preservation of detailed image information in RAW format facilitates enhanced feature extraction, thereby improving detection accuracy and resilience under low-light conditions. These findings provide novel insights into advancing object detection methodologies for autonomous driving systems operating in challenging lighting environments.

Research of anomaly detection based on dynamic anomaly detection enhancement framework

Abstract Anomaly detection plays a crucial role in various fields, from industrial defect inspection to geological detection. However, traditional approaches often struggle with insufficient discriminability and an inability to generalize to unseen anomalies. These limitations stem from the practical difficulty in gathering a comprehensive set of anomalies and the tendency to overlook anomalous instances in favor of normal samples. To address these challenges, we propose a novel Dynamic Anomaly Detection Enhancement Framework, integrating three key innovations: (1) SaliencyAug: An adaptive saliency-guided augmentation method that generates realistic pseudo-samples to enhance learning of rare anomalies, improving model generalization. (2) DynAB: A dynamic attention block that achieves effective multi-level feature fusion while minimizing redundant information, enhancing detection accuracy. (3) DualOM: A dual-head optimization module which employs separate heads for normal and anomalous sample learning, creating more explicit and discriminative decision boundaries. Extensive experiments across multiple real-world datasets demonstrate our framework's superior performance in detecting a wide range of anomalies, demonstrating 2.4% improvement over state-of-the-art methods.

Quantum intrusion detection system using outlier analysis

In the field of cybersecurity, hackers often enter computer systems despite current security measures, owing to the huge amount of network traffic that makes intruder identification difficult. Differentiating between authorized traffic and abnormal data produced by Distributed Denial of Service (DDoS) attackers is still a major difficulty. This research provides a unique technique that uses Quantum Machine Learning (QML) to improve security protocols for secure communication between two parties. Our strategy enhances detection accuracy and speed by using the characteristics of quantum neural networks. The approach includes creating a dataset from network traffic patterns, preprocessing it, and then turning it into quantum bits via angle embedding. The research uses outlier analysis, min-entropy, and quantum state fidelity to distinguish between normal and abnormal data patterns. The dispersed randomness of network header data is measured using entropy, which aids in identifying security concerns. The suggested QML-based methodology outperforms conventional approaches and current models, including AMM-CNN and ANN models, with a detection accuracy of 99.87% for DDoS attacks. This advancement leads to more effective and secure communication networks.

Enhancing Structured Query Language Injection Detection with Trustworthy Ensemble Learning and Boosting Models Using Local Explanation Techniques

This paper presents a comparative analysis of several decision models for detecting Structured Query Language (SQL) injection attacks, which remain one of the most prevalent and serious security threats to web applications. SQL injection enables attackers to exploit databases, gain unauthorized access, and manipulate data. Traditional detection methods often struggle due to the constantly evolving nature of these attacks, the increasing complexity of modern web applications, and the lack of transparency in the decision-making processes of machine learning models. To address these challenges, we evaluated the performance of various models, including decision tree, random forest, XGBoost, AdaBoost, Gradient Boosting Decision Tree (GBDT), and Histogram Gradient Boosting Decision Tree (HGBDT), using a comprehensive SQL injection dataset. The primary motivation behind our approach is to leverage the strengths of ensemble learning and boosting techniques to enhance detection accuracy and robustness against SQL injection attacks. By systematically comparing these models, we aim to identify the most effective algorithms for SQL injection detection systems. Our experiments show that decision tree, random forest, and AdaBoost achieved the highest performance, with an accuracy of 99.50% and an F1 score of 99.33%. Additionally, we applied SHapley Additive exPlanations (SHAPs) and Local Interpretable Model-agnostic Explanations (LIMEs) for local explainability, illustrating how each model classifies normal and attack cases. This transparency enhances the trustworthiness of our approach to detecting SQL injection attacks. These findings highlight the potential of ensemble methods to provide reliable and efficient solutions for detecting SQL injection attacks, thereby improving the security of web applications.

Enhanced federated anomaly detection through autoencoders using summary statistics-based thresholding

In Federated Learning, Anomaly Detection poses significant challenges due to the decentralized nature of data, especially under Non-IID distributions. This study proposes a federated threshold calculation method that aggregates summary statistics from normal and anomalous data across clients to create a global threshold for Anomaly Detection with federated Autoencoders, enhancing detection accuracy and robustness while ensuring privacy. Extensive experiments on datasets, including Credit Card Fraud Detection, Shuttle, and Covertype, show that our approach consistently outperforms existing federated and local threshold calculation methods. These findings highlight the potential of summary statistics in improving federated Anomaly Detection under Non-IID conditions.

A Machine Learning-Based Approach for Predicting the Level of Palm Oil Adulteration in Coconut Oil

Coconut oil, prized for its health benefits, faces quality threats from adulteration, particularly with cheaper palm oil. This not only degrades the quality but also poses health risks. Traditional detection methods are often labor-intensive, destructive, and time-consuming. This study addresses the issue by applying multispectral imaging technology combined with machine learning to detect palm oil adulteration in coconut oil. We selected four machine learning algorithms—Support Vector Machines (SVM), Quadratic Discriminant Analysis (QDA), K-Nearest Neighbors (KNN), and Bagging—due to their robustness in handling complex datasets. These models achieved classification accuracies of up to 100%, far surpassing traditional chemical tests, which are slower and dependent on tester expertise. To further enhance detection accuracy, we employed both hard- and soft-voting mechanisms, integrating the strengths of individual models to improve overall reliability. This research marks a significant advancement in detecting coconut oil adulteration, offering a faster, more efficient solution to ensure product quality and consumer health.

High-Precision and Lightweight Model for Rapid Safety Helmet Detection.

This paper presents significant improvements in the accuracy and computational efficiency of safety helmet detection within industrial environments through the optimization of the you only look once version 5 small (YOLOv5s) model structure and the enhancement of its loss function. We introduce the convolutional block attention module (CBAM) to bolster the model's sensitivity to key features, thereby enhancing detection accuracy. To address potential performance degradation issues associated with the complete intersection over union (CIoU) loss function in the original model, we implement the modified penalty-decay intersection over union (MPDIoU) loss function to achieve more stable and precise bounding box regression. Furthermore, considering the original YOLOv5s model's large parameter count, we adopt a lightweight design using the MobileNetV3 architecture and replace the original squeeze-and-excitation (SE) attention mechanism with CBAM, significantly reducing computational complexity. These improvements reduce the model's parameters from 15.7 GFLOPs to 5.7 GFLOPs while increasing the mean average precision (mAP) from 82.34% to 91.56%, demonstrating its superior performance and potential value in practical industrial applications.

The Detection and Classification of Scaphoid Fractures in Radiograph by Using a Convolutional Neural Network.

Scaphoid fractures, particularly occult and non-displaced fractures, are difficult to detect using traditional X-ray methods because of their subtle appearance and variability in bone density. This study proposes a two-stage CNN approach to detect and classify scaphoid fractures using anterior-posterior (AP) and lateral (LA) X-ray views for more accurate diagnosis. This study emphasizes the use of multi-view X-ray images (AP and LA views) to improve fracture detection and classification. The multi-view fusion module helps integrate information from both views to enhance detection accuracy, particularly for occult fractures that may not be visible in a single view. The proposed method includes two stages, which are stage 1: detect the scaphoid bone using Faster RCNN and a Feature Pyramid Network (FPN) for region proposal and small object detection. The detection accuracy for scaphoid localization is 100%, with Intersection over Union (IoU) scores of 0.8662 for AP views and 0.8478 for LA views. And stage 2: perform fracture classification using a ResNet backbone and FPN combined with a multi-view fusion module to combine features from both AP and LA views. This stage achieves a classification accuracy of 89.94%, recall of 87.33%, and precision of 90.36%. The proposed model performs well in both scaphoid bone detection and fracture classification. The multi-view fusion approach significantly improves recall and accuracy in detecting fractures compared to single-view approaches. In scaphoid detection, both AP and LA views achieved 100% detection accuracy. In fracture detection, using multi-view fusion, the accuracy for AP views reached 87.16%, and for LA views, it reached 83.83%. The multi-view fusion model effectively improves the detection of scaphoid fractures, particularly in cases of occult and non-displaced fractures. The model provides a reliable, automated approach to assist clinicians in detecting and diagnosing scaphoid fractures more efficiently.

HyMNet: A Multimodal Deep Learning System for Hypertension Prediction Using Fundus Images and Cardiometabolic Risk Factors

Study Objectives: This study aimed to develop a multimodal deep learning (MMDL) system called HyMNet, integrating fundus images and cardiometabolic factors (age and sex) to enhance hypertension (HTN) detection. Methods: HyMNet employed RETFound, a model pretrained on 1.6 million retinal images, for the fundus data, in conjunction with a fully connected neural network for age and sex. The two pathways were jointly trained by joining their feature vectors into a fusion network. The system was trained on 5016 retinal images from 1243 individuals provided by the Saudi Ministry of National Guard Health Affairs. The influence of diabetes on HTN detection was also assessed. Results: HyMNet surpassed the unimodal system, achieving an F1 score of 0.771 compared to 0.745 for the unimodal model. For diabetic patients, the F1 score was 0.796, while it was 0.466 for non-diabetic patients. Conclusions: HyMNet exhibited superior performance relative to unimodal approaches, with an F1 score of 0.771 for HyMNet compared to 0.752 for models trained on demographic data alone, underscoring the advantages of MMDL systems in HTN detection. The findings indicate that diabetes significantly impacts HTN prediction, enhancing detection accuracy among diabetic patients. Utilizing MMDL with diverse data sources could improve clinical applicability and generalization.

HADNet: A Novel Lightweight Approach for Abnormal Sound Detection on Highway Based on 1D Convolutional Neural Network and Multi-Head Self-Attention Mechanism

Video surveillance is an effective tool for traffic management and safety, but it may face challenges in extreme weather, low visibility, areas outside the monitoring field of view, or during nighttime conditions. Therefore, abnormal sound detection is used in traffic management and safety as an auxiliary tool to complement video surveillance. In this paper, a novel lightweight method for abnormal sound detection based on 1D CNN and Multi-Head Self-Attention Mechanism on the embedded system is proposed, which is named HADNet. First, 1D CNN is employed for local feature extraction, which minimizes information loss from the audio signal during time-frequency conversion and reduces computational complexity. Second, the proposed block based on Multi-Head Self-Attention Mechanism not only effectively mitigates the issue of disappearing gradients, but also enhances detection accuracy. Finally, the joint loss function is employed to detect abnormal audio. This choice helps address issues related to unbalanced training data and class overlap, thereby improving model performance on imbalanced datasets. The proposed HADNet method was evaluated on the MIVIA Road Events and UrbanSound8K datasets. The results demonstrate that the proposed method for abnormal audio detection on embedded systems achieves high accuracy of 99.6% and an efficient detection time of 0.06 s. This approach proves to be robust and suitable for practical applications in traffic management and safety. By addressing the challenges posed by traditional video surveillance methods, HADNet offers a valuable and complementary solution for enhancing safety measures in diverse traffic conditions.

FEI-YOLO: A Lightweight Soybean Pod-Type Detection Model

Identifying and statistically analyzing soybean pod types are crucial for seed evaluation and yield estimation. Traditional visual assessment by breeding personnel is time-consuming, labor-intensive, and prone to subjective bias, especially with large datasets. Automatic assessment methods usually struggle with the highly confusing pod types with two and three seeds, affecting the model’s identification accuracy. To address these issues, we propose to improve the standard YOLOv5s object detection model to enhance the differentiation between pod types and to boost the model’s efficiency in prediction. To reduce the number of parameters and the computational load, we propose to introduce the FasterNet Block module in the FasterNet model into the original C3 module, leading to improvements in both detection accuracy and speed. To strengthen the feature extraction and representation for specific targets, the Efficient Multi-Scale Attention (EMA) module is incorporated into the C3 module of the backbone network, improving the identification of similar pod types. The Inner-IoU is combined with the CIoU as the loss function to further enhance detection accuracy and generalization. Experiments comparing FEI-YOLO with the baseline YOLOv5s show that FEI-YOLO achieves an mAP@0.5 of 98.6% and an mAP@0.5:0.95 of 81.1%, with improvements of 1.5% and 1.4%, respectively. Meanwhile, the number of parameters is reduced by 13.2%, and FLOPs decreased by 10.8%, demonstrating the model’s effectiveness and efficiency, enabling rapid and accurate identification of soybean pod types from images.

Using U-Net models in deep learning for brain tumor detection from MRI scans

Tumor diseases in the nervous system are both dangerous and complex. Magnetic Resonance Imaging (MRI) is crucial for detecting brain disease; however, identifying the presence of tumors from these is time-consuming and requires a professional doctor. Utilizing deep learning for tumor detection in MRI images can reduce waiting times and enhance detection accuracy. We propose a method employing two U-Net models: ResNeXt-50 and EfficientNet architectures, integrated with a Feature Pyramid Network (FPN) for segmenting brain tumor. The models were trained on the BraTS 2021 dataset, consisting of 3,929 MRI scan images with 3,929 corresponding masks, divided into training, testing, and evaluation sets in a 70:15:15 ratio. The results indicate that the hybrid model, which combines EfficientNet and FPN, delivers superior performance, with an average Intersection over Union (IoU) accuracy of 0.90 on the test set compared to 0.50 for ResNeXt-50, and Dice accuracy of 0.92 compared to 0.66 for ResNeXt-50. Furthermore, we developed a web application that implements the EfficientNet with FPN model, facilitating convenient tumor detection from uploaded MRI images for doctors.