Detection Model Research Articles

AI-IoT based smart agriculture pivot for plant diseases detection and treatment

There are some key problems faced in modern agriculture that IoT-based smart farming. These problems such shortage of water, plant diseases, and pest attacks. Thus, artificial intelligence (AI) technology cooperates with the Internet of Things (IoT) toward developing the agriculture use cases and transforming the agriculture industry into robustness and ecologically conscious. Various IoT smart agriculture techniques are escalated in this field to solve these challenges such as drop irrigation, plant diseases detection, and pest detection. Several agriculture devices were installed to perform these techniques on the agriculture field such as drones and robotics but in expense of their limitations. This paper proposes an AI-IoT smart agriculture pivot as a good candidate for the plant diseases detection and treatment without the limitations of both drones and robotics. Thus, it presents a new IoT system architecture and a hardware pilot based on the existing central pivot to develop deep learning (DL) models for plant diseases detection across multiple crops and controlling their actuators for the plant diseases treatment. For the plant diseases detection, the paper augments a dataset of 25,940 images to classify 11-classes of plant leaves using a pre-trained ResNet50 model, which scores the testing accuracy of 99.8%, compared to other traditional works. Experimentally, the F1-score, Recall, and Precision, for ResNet50 model were 99.91%, 99.92%, and 100%, respectively.

Automated Lumbar Spine Object Detection

Accurate analysis of lumbar spine X-ray images is critical for early diagnosis and treatment of spinal conditions. This research leverages YOLOv8, a state-of-the-art object detection model, to automate the detection and localization of lumbar spine abnormalities. By fine-tuning YOLOv8 on a custom lumbar spine dataset, the model achieves high detection accuracy with real-time performance. The methodology includes data augmentation, model training, evaluation, and visualization of results. This study demonstrates the effectiveness of deep learning in enhancing radiological workflows and improving diagnostic accuracy in medical imaging.A custom-labelled lumbar spine X-ray dataset was curated and used to fine-tune the YOLOv8 model. The methodology involved rigorous data preprocessing, targeted data augmentation strategies, model training under transfer learning protocols, and comprehensive evaluation using standard object detection metrics such as Precision, Recall, and mean Average Precision (mAP). The results demonstrate that the fine-tuned model achieves a mAP50 score of 91.5%, with a precision of 92% and recall of 89%, outperforming conventional CNN-based approaches in both detection accuracy and speed. Visualizations of the model's outputs confirm its ability to accurately localize vertebrae, identify degenerative changes, and flag abnormal regions of interest. These findings underscore the potential of YOLOv8 as a reliable and efficient tool to assist radiologists, enhance clinical workflows, and improve diagnostic accuracy, particularly in resource-constrained healthcare environments. Future work will explore integration with segmentation models and broader validation across diverse patient demographics to further solidify the system's clinical applicability.

Efficient polyp detection algorithm based on deep learning

Objective Colon polyp detection is crucial in reducing the incidence of colorectal cancer. However, due to the diverse morphology of colon polyps, their high similarity to surrounding tissues, and the difficulty of detecting small target polyps, false negatives and false positives are common problems. Methods To address this, we propose a lightweight and efficient colon polyp detection model based on YOLOv10, a deep learning-based object detection method—EP-YOLO (Efficient for Polyp). By introducing the GBottleneck module, we reduce the number of parameters and accelerate inference; a lightweight GHead detection head and an additional small target detection layer are designed to enhance small target recognition ability; we propose the SE_SPPF module to improve attention on polyps while suppressing background noise interference; the loss function is replaced with Wise-IoU to optimize gradient distribution and improve generalization ability. Results Experimental results on the publicly available LDPolypVideo (7,681 images), Kvasir-SEG (1,000 images) and CVC-ClinicDB (612 images) datasets show that EP-YOLO achieves precision scores of 94.17%, 94.32% and 93.21%, respectively, representing improvements of 2.10%, 2.05% and 1.42% over the baseline algorithm, while reducing the number of parameters by 16%. Conclusion Compared with other mainstream object detection methods, EP-YOLO demonstrates significant advantages in accuracy, computational load and FPS, making it more suitable for practical medical scenarios in colon polyp detection.

DEL_YOLO: A Lightweight Coal-Gangue Detection Model for Limited Equipment

DEL_YOLO: A Lightweight Coal-Gangue Detection Model for Limited Equipment

Improved SSS small target detection method based on kernel regression and patch-image model

Abstract With the rapid advancement of unmanned technology, autonomous underwater vehicles equipped with side-scan sonar are playing an increasingly vital role in the realm of underwater exploration. The detection of underwater small targets such as mine-like objects, unexploded ordnance, and rod-shaped items is a focal point of current sonar technology research, playing a crucial role in military struggle. However, existing methods overly rely on the prior shadow information and are prone to missing numerous small targets. To address this, we propose a weighted SSS patch-image model. The method is an improved method for small target detection in SSS images based on iterative steering kernel regression and patch-image model. The method enables effective detection of small targets without considering shadow information. Firstly, kernel regression is employed using steering kernels to denoise the images while preserving edge information. Subsequently, a small target detection model is constructed using the patch-image model by considering the reconstructed SSS image, the target image, the background image, and the noise image. According to the experimental results, the improved small target detection method combining the two aforementioned algorithms demonstrates high accuracy and reliability in detecting underwater small targets on SSS images. Comparative experiments further reveal that this approach overcomes the limitations of traditional methods that rely heavily on target shadow information, establishing it as an efficient and robust solution for underwater small target detection in SSS images.

A preprocessing toolbox for 2-photon subcellular calcium imaging.

Recording the spiking activity from subcellular compartments of neurons such as axons and dendrites during mouse behavior with 2-photon calcium imaging is increasingly common yet remains challenging due to low signal-to-noise, inaccurate region-of-interest (ROI) identification, movement artifacts, and difficulty in grouping ROIs from the same neuron. To address these issues, we present a computationally efficient pre-processing pipeline for subcellular signal detection, movement artifact identification, and ROI grouping. For subcellular signal detection, we capture the frequency profile of calcium transient dynamics by applying Fast Fourier Transform (FFT) on smoothed time-series calcium traces collected from axon ROIs. We then apply band-pass filtering methods (e.g. 0.05 to 0.12 Hz) to select ROIs that contain frequencies that match the power band of transients. To remove motion artifacts from z-plane movement, we apply Principal Component Analysis on all calcium traces and use a Bottom-Up Segmentation change-point detection model on the first principal component. After removing movement artifacts, we further identify calcium transients from noise by analyzing their prominence and duration. Finally, ROIs with high activity correlation are grouped using hierarchical or k-means clustering. Using axon ROIs in the CA1 region, we confirm that both clustering methods effectively determine the optimal number of clusters in pairwise correlation matrices, yielding similar groupings to "ground truth" data. Our approach provides a guideline for standardizing the extraction of physiological signals from subcellular compartments during rodent behavior with 2-photon calcium imaging.Significance Statement The SUBPREP pipeline is specifically designed to process calcium imaging data from axons, dendrites, and other subcellular structures, which pose unique challenges due to their low signal-to-noise ratios, susceptibility to movement artifacts, and complex morphologies. This pipeline enables researchers to extract reliable physiological signals from noisy datasets, making it a significant improvement over manual or neural network-based approaches. By addressing key bottlenecks in subcellular imaging, this toolbox facilitates the standardization of preprocessing workflows, which is critical for reproducibility and cross-study comparisons. These improvements are especially timely as subcellular calcium imaging becomes increasingly common in neuroscience research. SUBPREP's potential for broad adoption, coupled with its ability to uncover profound biological insights, makes it a valuable contribution to in vivo imaging.

GESC-YOLO: Improved Lightweight Printed Circuit Board Defect Detection Based Algorithm

Printed circuit boards (PCBs) are an indispensable part of electronic products, and their quality is crucial to the operational integrity and functional reliability of these products. Currently, existing PCB defect detection models are beset with issues such as excessive model size and parameter complexity, rendering them ill-equipped to meet the requirements for lightweight deployment on mobile devices. To address this challenge, this paper proposes a lightweight detection model, GESC-YOLO, developed through modifications to the YOLOv8n architecture. First, a new lightweight module, C2f-GE, is designed to replace the C2f module of the backbone network, which effectively reduces the computational parameters, and at the same time increases the number of channels of the feature map to enhance the feature extraction capability of the model. Second, the neck network employs the lightweight hybrid convolution GSConv. By integrating it with the VoV-GSCSP module, the Slim-neck structure is constructed. This approach not only guarantees detection precision but also enables model lightweighting and a reduction in the number of parameters. Finally, the coordinate attention is introduced into the neck network to decompose the channel attention and aggregate the features, which can effectively retain the spatial information and thus improve the detection and localization accuracy of tiny defects (defect area less than 1% of total image area) in PCB defect images. Experimental results demonstrate that, in contrast to the original YOLOv8n model, the GESC-YOLO algorithm boosts the mean Average Precision (mAP) of PCB surface defects by 0.4%, reaching 99%. Simultaneously, the model size is reduced by 25.4%, the parameter count is cut down by 28.6%, and the computational resource consumption is reduced by 26.8%. This successfully achieves the harmonization of detection precision and model lightweighting.

Multi-centric AI Model for Unruptured Intracranial Aneurysm Detection and Volumetric Segmentation in 3D TOF-MRI.

The aim of this study was to develop an open-source nnU-Net-based AI model for combined detection and segmentation of unruptured intracranial aneurysms (UICA) in 3D TOF-MRI and compare models trained on datasets with aneurysm-like differential diagnoses. This retrospective study (2020-2023) included 385 anonymized 3D TOF-MRI images from 345 patients (mean age 59years, 60% female) at multiple centers plus 113 subjects from the ADAM challenge. Images featured untreated or possible UICA and differential diagnoses. Four distinct training datasets were created, and the nnU-Net framework was used for model development. Performance was assessed on a separate test set using sensitivity and false positive (FP)/case rate for detection and DICE score and NSD (normalized surface distance, 0.5mm threshold) for segmentation. Segmentation performance on the test set was also compared to a second human reader. The four models achieved overall sensitivity between 82 and 85% and an FP/case rate of 0.20 to 0.31, with no significant differences (p = 0.90 and p = 0.16) between them. The primary model showed 85% sensitivity and 0.23 FP/case rate, outperforming the ADAM-challenge winner (61%) and a nnU-Net trained on ADAM data (51%) in sensitivity (p < 0.05). Mean DICE (0.73) and NSD (0.84 for 0.5mm threshold) for correctly detected UICA did not significantly differ from human reader performance. Our open-source, nnU-Net-based AI model (available at https://zenodo.org/records/13386859 ) demonstrates high sensitivity, low FP rates, and consistent segmentation accuracy for UICA detection and segmentation in 3D TOF-MRI, suggesting its potential to improve clinical diagnosis and monitoring of UICA.

The Potential Role of AI- and Machine Learning Models in the Early Detection of Oral Cancer and Oral Potentially Malignant Disorders.

Artificial Intelligence (AI) is playing an increasing role in advancing diagnostic processes and decision-making in healthcare. In the early detection of oral cancer and oral potentially malignant disorders (OPMDs), its role is still being explored. This paper evaluates advancements in AI applications for the early detection of oral cancer and OPMDs. A narrative umbrella review was performed on reviews that explicitly evaluated non-invasive diagnostic techniques combined with AI-modalities or machine learning techniques in the early detection of oral cancer and OPMDs. Key findings of eight studies published between 2015 and 2024 demonstrate various AI-modalities and their diagnostic accuracy, accessibility and affordability, limitations and challenges and ethical and regulatory needs. AI- and deep learning models hold promise in improving the early detection of oral cancer and OPMDs, offering high diagnostic accuracy that can significantly enhance patient outcomes. Challenges such as limited explainability and ethical concerns must be addressed to fully integrate these technologies into daily clinical practice.

CASEMark: A hybrid model for robust anatomical landmark detection in multi-structure X-rays

Anatomical landmark detection is crucial in medical image analysis, facilitating accurate diagnosis, surgical planning, and treatment evaluation. However, existing methods often struggle to simultaneously capture global context and local details while exhibiting limited generalization across diverse datasets and imaging modalities. To relieve this, we propose a hybrid model that leverages convolutional operations to capture local information and a Swin Transformer to enhance global context. Specifically, we introduce a novel U-shaped architecture, termed Convolutional Attention Swin Enhanced Landmark Detection Network (CASEMark). CASEMark integrates three key innovations: (1) a Convolutional Attention Swin Transformer module (CAST) that integrates transformer-based global context modeling with convolutional operations for local feature extraction, (2) an Enhanced Skip Attention Module (ESAM) enabling adaptive feature fusion between encoder and decoder pathways, and (3) a multi-resolution heatmap learning strategy that aggregates information across scales. This approach effectively balances global-local feature extraction with robust cross-modality generalization. Extensive experiments on four public datasets demonstrate the superiority of CASEMark. The code and datasets will be made publicly available.

Energy-Efficient Fall-Detection System Using LoRa and Hybrid Algorithms

Wearable fall-detection systems have received significant research attention during the last years. Fall detection in wearable devices presents key challenges, particularly in balancing high precision with low power consumption—both of which are essential for the continuous monitoring of older adults and individuals with reduced mobility. This study introduces a hybrid system that integrates a threshold-based model for preliminary detection with a deep learning-based approach that combines a CNN (Convolutional Neural Network) for spatial feature extraction with a LSTM (Long Short-Term Memory) model for temporal pattern recognition, aimed at improving classification accuracy. LoRa technology enables long-range, energy-efficient communication, ensuring real-time monitoring across diverse environments. The wearable device operates in ultra-low-power mode, capturing acceleration data at 20 Hz and transmitting a 4-s window when a predefined threshold in the acceleration magnitude is exceeded. The CNN-LSTM classifier refines event identification, significantly reducing false positives. This design extends operational autonomy to 178 h of continuous monitoring. The experimental and systematic evaluation of the prototype achieved a 96.67% detection rate (sensitivity) for simulated falls and a 100% specificity in classifying conventional Activities of Daily Living as non-falls. These results establish the system as a robust and scalable solution, effectively addressing limitations in power efficiency, connectivity, and detection accuracy while enhancing user safety and quality of life.

Detecting Activities of Daily Living in Egocentric Video to Contextualize Hand Use at Home in Outpatient Neurorehabilitation Settings.

Wearable egocentric cameras and machine learning have the potential to provide clinicians with a more nuanced understanding of patient hand use at home after stroke and spinal cord injury (SCI). However, they require detailed contextual information (i.e., activities and object interactions) to effectively interpret metrics and meaningfully guide therapy planning. We demonstrate that an object-centric approach, focusing on what objects patients interact with rather than how they move, can effectively recognize Activities of Daily Living (ADL) in real-world rehabilitation settings. We evaluated our models on a complex dataset collected in the wild comprising 2261 minutes of egocentric video from 16 participants with impaired hand function. By leveraging pre-trained object detection and hand-object interaction models, our system achieves robust performance across different impairment levels and environments, with our best model achieving a mean weighted F1-score of 0.78 ± 0.12 and maintaining an F1-score over 0.5 for all participants using leave-one-subject-out cross validation. Through qualitative analysis, we observe that this approach generates clinically interpretable information about functional object use while being robust to patient-specific movement variations, making it particularly suitable for rehabilitation contexts with prevalent upper limb impairment.

Driver Monitoring System

Abstract The Driver Monitoring System (DMS) is an innovative solution aimed at reducing road accidents caused by driver fatigue, distraction, and other unsafe behaviors. Leveraging advanced technologies like Convolutional Neural Networks (CNNs) and YOLO object detection models, the system monitors critical parameters such as eye movements, blinking frequency, and facial expressions. This real-time analysis enables the detection of drowsiness, distractions, and risky behaviors, issuing timely alerts to ensure driver safety. CNNs, with their superior feature extraction and classification capabilities, outperform traditional methods by offering high accuracy, robustness to variations, and scalability. This project demonstrates the efficacy of integrating CNNs into DMS, paving the way for safer driving environments and significant advancements in automotive safety.

Enhanced Propaganda Detection in Public Social Media Discussions Using a Fine-Tuned Deep Learning Model: A Diffusion of Innovation Perspective

During the COVID-19 pandemic, social media platforms emerged as both vital information sources and conduits for the rapid spread of propaganda and misinformation. However, existing studies often rely on single-label classification, lack contextual sensitivity, or use models that struggle to effectively capture nuanced propaganda cues across multiple categories. These limitations hinder the development of robust, generalizable detection systems in dynamic online environments. In this study, we propose a novel deep learning (DL) framework grounded in fine-tuning the RoBERTa model for a multi-label, multi-class (ML-MC) classification task, selecting RoBERTa due to its strong contextual representation capabilities and demonstrated superiority in complex NLP tasks. Our approach is rigorously benchmarked against traditional and neural methods, including, TF-IDF with n-grams, Conditional Random Fields (CRFs), and long short-term memory (LSTM) networks. While LSTM models show strong performance in capturing sequential patterns, our RoBERTa-based model achieves the highest overall accuracy at 88%, outperforming state-of-the-art baselines. Framed within the diffusion of innovations theory, the proposed model offers clear relative advantages—including accuracy, scalability, and contextual adaptability—that support its early adoption by Information Systems researchers and practitioners. This study not only contributes a high-performing detection model but also delivers methodological and theoretical insights for combating propaganda in digital discourse, enhancing resilience in online information ecosystems.

Bearing ring end face defect detection based on improved YOLOv5s

Abstract Bearings, extensively utilized in the industrial sector, play a pivotal role in the defect detection of industrial components. This paper proposes a defect detection algorithm based on an enhanced YOLOv5s to improve the accuracy and speed of bearing ring end face defect detection. The algorithm boosts minimal target detection by incorporating a small object detection head and combining multi-scale representation learning with anchor box calculations. It introduces a formula to calculate the receptive field sizes of convolutional layers. By calculating the receptive field sizes corresponding to the feature map pixels and integrating this data with anchor analysis, three anchors optimal for the small object detection head are determined. Additionally, an attention mechanism refines the receptive fields of the neural network's output feature maps, enhancing the model's performance. Extensive experiments on a dataset of bearing ring end face defects from industrial sites reveal that the improved YOLOv5s algorithm achieves a detection accuracy (mAP) of 96.14%, a detection speed of up to 44 FPS, and a model size of 20.08M. Compared to other mainstream detection models, this algorithm not only meets but exceeds the real-time detection requirements of industrial production in terms of accuracy and model complexity. With its high precision, compact model size, and rapid detection speed, this algorithm provides a robust foundation for quality control in bearing production.&amp;#xD;

Harnessing Semantic and Trajectory Analysis for Real-Time Pedestrian Panic Detection in Crowded Micro-Road Networks

Pedestrian panic behavior is a primary cause of overcrowding and stampede accidents in public micro-road network areas with high pedestrian density. However, reliably detecting such behaviors remains challenging due to their inherent complexity, variability, and stochastic nature. Current detection models often rely on single-modality features, which limits their effectiveness in complex and dynamic crowd scenarios. To overcome these limitations, this study proposes a contour-driven multimodal framework that first employs a CNN (CDNet) to estimate density maps and, by analyzing steep contour gradients, automatically delineates a candidate panic zone. Within these potential panic zones, pedestrian trajectories are analyzed through LSTM networks to capture irregular movements, such as counterflow and nonlinear wandering behaviors. Concurrently, semantic recognition based on Transformer models is utilized to identify verbal distress cues extracted through Baidu AI’s real-time speech-to-text conversion. The three embeddings are fused through a lightweight attention-enhanced MLP, enabling end-to-end inference at 40 FPS on a single GPU. To evaluate branch robustness under streaming conditions, the UCF Crowd dataset (150 videos without panic labels) is processed frame-by-frame at 25 FPS solely for density assessment, whereas full panic detection is validated on 30 real Itaewon-Stampede videos and 160 SUMO/Unity simulated emergencies that include explicit panic annotations. The proposed system achieves 91.7% accuracy and 88.2% F1 on the Itaewon set, outperforming all single- or dual-modality baselines and offering a deployable solution for proactive crowd safety monitoring in transport hubs, festivals, and other high-risk venues.

Memory optimized random forest classifier for EEG seizure detection in implantable monitoring and closed-loop neurostimulation devices.

Objective&#xD;Up to one third of epilepsy patients do not achieve satisfactory seizure control and may benefit from implantable devices for responsive neurostimulation or online seizure monitoring. Beyond energy efficiency, the limited memory capacity in these devices, imposes significant constraints to algorithmic design of seizure detection models. This study aims to evaluate the performance of cross-patient random forest (RF) models optimized for low-power microcontroller applications by assessing various channel integration strategies and measuring their memory requirements. &#xD;&#xD;Approach&#xD;Fifty patients undergoing electroencephalographic monitoring with 362 seizures were included in the analysis, with approximately one hour of signal for each seizure. One central and four peripheral electrodes over the epileptogenic focus were selected to resemble the layout of a novel neurostimulation device. Fifteen features were extracted from 2-second non-overlapping segments. RF models comprised either 500 or 125 trees, with varying depths. Three early channel integration (EI) strategies were compared with late integration (LI), using three channel fusion methods. A leave-one-patient-out cross-validation approach was used for evaluation, and memory requirements, alongside with inference energy and latency for 8-bit integer and 32-bit floating point models were computed on a microcontroller. &#xD;&#xD;Main results&#xD;The performance of EI feature sorting and LI were comparable. LI was favored by the 32-bit floating point format and more complex models, with the median channel fusion achieving a median ROC-AUC score of 0.925. Feature sorting performed best with medium-sized models and was largely unaffected by the 8-bit integer format. Following causal output post-processing, false stimulations per hour were reduced to 5.5 at 100% sensitivity and fell below 3 at ~80% sensitivity. &#xD;&#xD;Significance&#xD;Our findings suggest that RF models with minimal energy and memory requirements can achieve state-of-the-art performance, making them well-suited for embedded applications in implantable devices. The complex interplay of the investigated factors is critical to performance, and along hardware specifications, should guide algorithmic design. &#xD.

MAS-YOLO: a fabric defect detection network based on YOLOv8

Defect detection is crucial for controlling product quality in the process of textile production. However, for existing detection techniques, there are still challenges in identifying different forms of defect and small defects within the same category. To address this issue, we propose a fabric defect detection model called MAS-YOLO. This model is based on YOLOv8n and incorporates several key innovations. First, we designed a multi-branch coordinate attention module to capture direction and position information. Second, we designed an adaptive weighted downsampling module based on grouped convolution, which emphasizes defective features and reduces background interference using weighting features. Finally, we introduced sliding loss to address the imbalance between easy and difficult samples. The experimental results show that the mean average precisions for a customized fabric defect dataset and the AliCloud Tianchi dataset were 96.3% and 51.6%, respectively, that is, 6.9% and 7.8% higher, respectively, than the original YOLOv8n. The detection speeds using the GTX1050ti graphics card and RTX3070ti graphics card are 57.3 frames per second (fps) and 154.3 fps, respectively; this can meet the real-time requirements of defect detection in most industrial sites and provide technical support for the application of lightweight network models in the industry.

InDepth: A Distributed Data Collection System for Modern Computer Networks

Cybersecurity researchers and security analysts rely heavily on data to train and test network threat detection models, and to conduct post-breach forensic analyses. Comprehensive data-including network traces, host telemetry, and contextual information-are crucial for these tasks. However, widely used public datasets often suffer from outdated network traffic and features, statistical anomalies, and simulation artifacts. Furthermore, existing data collection systems frequently face architectural and computational limitations, necessitating workarounds that result in incomplete or disconnected data. Currently, no framework provides comprehensive data collection from all network segments without requiring specialized or proprietary hardware or software agents. This paper introduces InDepth, a scalable system employing a distributed, data-link layer architecture that enables comprehensive data acquisition across entire networks. We also present a model cyber range capable of dynamically generating datasets for evaluation. We demonstrate the effectiveness of InDepth using real-world network data.

YOLO-TARC: YOLOv10 with Token Attention and Residual Convolution for Small Void Detection in Root Canal X-Ray Images

The detection of small voids or defects in X-ray images of tooth root canals still faces challenges. To address the issue, this paper proposes an improved YOLOv10 that combines Token Attention with Residual Convolution (ResConv), termed YOLO-TARC. To overcome the limitations of existing deep learning models in effectively retaining key features of small objects and their insufficient focusing capabilities, we introduce three improvements. First, ResConv is designed to ensure the transmission of discriminative features of small objects during feature propagation, leveraging the ability of residual connections to transmit information from one layer to the next. Second, to tackle the issue of weak focusing capabilities on small targets, a Token Attention module is introduced before the third small object detection head. By tokenizing feature maps and enhancing local focusing, it enables the model to pay closer attention to small targets. Additionally, to optimize the training process, a bounding box loss function is adopted to achieve faster and more accurate bounding box predictions. YOLO-TARC simultaneously enhances the ability to retain detailed information of small targets and improves their focusing capabilities, thereby increasing detection accuracy. Experimental results on a private root canal X-ray image dataset demonstrate that YOLO-TARC outperforms other state-of-the-art object detection models, achieving a 7.5% improvement to 80.8% in mAP50 and a 6.2% increase to 80.0% in Recall. YOLO-TARC can contribute to more accurate and efficient objective postoperative evaluation of root canal treatments.