Average Precision Research Articles

Query Expansion Using Proposed Location-Based Algorithm for Hindi–English CLIR: Analyzing Three Test Collections

The rapid growth of contents on the Web in different languages increases the demand of Cross-Lingual Information Retrieval (CLIR). The accuracy of result suffers due to many problems such as ambiguity and drift issue in query. Query Expansion (QE) offers reliable solution for obtaining suitable documents for user queries. In this paper, we proposed an architecture for Hindi–English CLIR system using QE for improving the relevancy of retrieved results. In this architecture, for the addition of term(s) at appropriate position(s), we proposed a location-based algorithm to resolve the drift query issue in QE. User queries in Hindi language have been translated into document language (i.e. English) and the accuracy of translation is improved using Back-Translation. Google search has been performed and the retrieved documents are ranked using Okapi BM25 to arrange the documents in the order of decreasing relevancy to select the most suitable terms for QE. We used term selection value (TSV) for QE and for retrieving the terms, we created three test collections namely the (i) description and narration of the Forum for Information Retrieval Evaluation (FIRE) dataset, (ii) Snippets of retrieved documents against each query and (iii) Nearest-Neighborhood (NN) words against each query word among the ranked documents. To evaluate the system, 50 queries of Hindi language are selected from the FIRE-2012 dataset. In this paper, we performed two experiments: (i) impact of the proposed location-based algorithm on the proposed architecture of CLIR; and (ii) analysis of QE using three datasets, i.e. FIRE, NN and Snippets. In the first case, result shows that the relevancy of Hindi–English CLIR is improved by performing QE using the location-based algorithm and a 12% of improvement is achieved as compared to the results of QE obtained without applying the location-based algorithm. In the second case, the location-based algorithm is applied on three datasets. The Mean Average Precision (MAP) values of retrieved documents after QE are 0.5379 (NN), 0.6018 (FIRE) and 0.6406 (Snippets) for the three test collections, whereas the MAP before QE is 0.37102. This clearly shows the significant improvement of retrieved results for all three test collections. Among the three test collections, QE has been found most effective along with Snippets as indicated by the results with the improvements of 6.48% and 19.12% over FIRE and NN test collections, respectively.

Visual Reinforcement Learning for Dynamic Object Detection

Abstract. Object detection is a widely studied task in computer vision. Current methods often focus on images captured from appropriate viewpoints. However, there is a large disparity between objects observed from different viewpoints in the real world. Dynamic Object Detection (DOD) method automatically adjusts the camera viewpoint in a visual scene to sequentially find optimal viewpoints. Currently, the DOD tasks are usually modeled as a sequential decision-making problem and solved using reinforcement learning methods. Existing approaches face challenges with sparse rewards and training instability. To tackle these issues, we proposed a single-step reward function and a lightweight network, respectively. The single-step reward function, which provides timely feedback, gives an efficient training process for DOD tasks. The lightweight network with few parameters can ensure the stability of the training process. To evaluate the effectiveness of our method, we developed a simulation dataset based on UE4, which consists of 1800 training images and 450 testing images. The dataset includes five object categories: vans, cars, trailers, box trucks and SUVs. Experiments demonstrate that our method outperforms SOTA object detectors on our simulation dataset. Specifically, the average precisions(APs) are improved from 89.1% to 96.0% when using the YOLOv8 object detector.

UAV and Deep Learning: Detection of selected riparian species along the Ganga River

Abstract. Environmental protection and sustainable natural resource management are being recognised worldwide as essential goals to safe guard human health and well-being. Riparian zones, that face the highest decline in freshwater biodiversity, are of prime conservation priority because they are essential for regulating climate, preserving aquatic-terrestrial biodiversity, maintaining ground water recharge and restoring rivers. In today's fast-paced data-driven environment, artificial intelligence (AI) is the precise answer to a wide range of problems including biodiversity conservation and wildlife management. Leveraging advancements like Uncrewed/Unmanned Aerial Vehicles (UAVs) and AI has resulted innovative strides in wildlife conservation. This study utilised UAV imagery to record high-resolution data of aquatic habitat and species along the Ganga River and employed deep learning algorithms to analyse the data. Through extensive field surveys in the Hastinapur Wildlife Sanctuary, 7,025 photos representing a variety of environments, including 20,000 annotated samples of aquatic animals such as turtles and gharials were generated. Vision based computing capabilities such as pattern recognition model were developed to identify these species. To enrich and enhance the dataset and model, we used different image pre-processing techniques. Slight rotation (±5 degrees), minor cropping (up to 10%), and adjustments in brightness, saturation, and shear (±15%) were applied. Controlled blur (up to 0.5%) and exposure modifications (±5%) were also implemented on the image dataset to improve accuracy. Three Convolutional Neural Network (CNN) architectures, single-stage detectors named YOLO v7, YOLO v8, and Roboflow 3.0, were used for detecting the select species. Results show YOLO v8 excels, achieving mean average precision (mAP) of 98.8% for gharial and 92.2% for turtle detection, with a rapid average detection time of 0.308 seconds per frame at 3200 × 3200 resolution. Additionally, our model demonstrates real-time species detection capability through innovative frame sampling techniques with UAVs. This methodology provides promising technique to collect scientific data on IUCN red listed turtles and critically endangered gharials, allowing detection, monitoring, and real time counting with minimal intrusion. In conclusion, the fusion of UAVs and deep learning promises to revolutionize habitat monitoring, aiding conservation decision-making.

Real-time detection of surface cracking defects for large-sized stamped parts

This study presents a framework for the real-time detection of surface cracking in large-sized stamped metal parts. The framework aims to address the challenges of low detection efficiency and high error rates associated with manual cracking detection. Within this framework, a novel network, SNF-YOLOv8, is proposed to efficiently detect cracking while ensuring that the detection speed matches the production speed. The network incorporates a convolutional spatial-to-depth module to enhance the detection of small-sized cracking and mitigate surface interference during inspections. Furthermore, a visual self-attention mechanism is introduced to improve feature extraction. A combination of standard convolutional and depth-wise separable convolutional layers in the neck network enhances speed without compromising accuracy. Experimental validation conducted using a dataset from actual production lines, in collaboration with a multi-national corporation, demonstrates that SNF-YOLOv8 achieves an average precision of 85.2% at a detection speed of 164 frames per second. The framework achieves an accuracy rate of 98.8% in detecting large-sized cracking and 96.4% in detecting small-sized cracking, meeting the requirements for high-precision and real-time detection applications.

EyeXNet: Enhancing Abnormality Detection and Diagnosis via Eye-Tracking and X-ray Fusion

Integrating eye gaze data with chest X-ray images in deep learning (DL) has led to contradictory conclusions in the literature. Some authors assert that eye gaze data can enhance prediction accuracy, while others consider eye tracking irrelevant for predictive tasks. We argue that this disagreement lies in how researchers process eye-tracking data as most remain agnostic to the human component and apply the data directly to DL models without proper preprocessing. We present EyeXNet, a multimodal DL architecture that combines images and radiologists’ fixation masks to predict abnormality locations in chest X-rays. We focus on fixation maps during reporting moments as radiologists are more likely to focus on regions with abnormalities and provide more targeted regions to the predictive models. Our analysis compares radiologist fixations in both silent and reporting moments, revealing that more targeted and focused fixations occur during reporting. Our results show that integrating the fixation masks in a multimodal DL architecture outperformed the baseline model in five out of eight experiments regarding average Recall and six out of eight regarding average Precision. Incorporating fixation masks representing radiologists’ classification patterns in a multimodal DL architecture benefits lesion detection in chest X-ray (CXR) images, particularly when there is a strong correlation between fixation masks and generated proposal regions. This highlights the potential of leveraging fixation masks to enhance multimodal DL architectures for CXR image analysis. This work represents a first step towards human-centered DL, moving away from traditional data-driven and human-agnostic approaches.

Research on Bearing Surface Scratch Detection Based on Improved YOLOV5

Bearings are crucial components of machinery and equipment, and it is essential to inspect them thoroughly to ensure a high pass rate. Currently, bearing scratch detection is primarily carried out manually, which cannot meet industrial demands. This study presents research on the detection of bearing surface scratches. An improved YOLOV5 network, named YOLOV5-CDG, is proposed for detecting bearing surface defects using scratch images as targets. The YOLOV5-CDG model is based on the YOLOV5 network model with the addition of a Coordinate Attention (CA) mechanism module, fusion of Deformable Convolutional Networks (DCNs), and a combination with the GhostNet lightweight network. To achieve bearing surface scratch detection, a machine vision-based bearing surface scratch sensor system is established, and a self-made bearing surface scratch dataset is produced as the basis. The scratch detection final Average Precision (AP) value is 97%, which is 3.4% higher than that of YOLOV5. Additionally, the model has an accuracy of 99.46% for detecting defective and qualified products. The average detection time per image is 263.4 ms on the CPU device and 12.2 ms on the GPU device, demonstrating excellent performance in terms of both speed and accuracy. Furthermore, this study analyzes and compares the detection results of various models, demonstrating that the proposed method satisfies the requirements for detecting scratches on bearing surfaces in industrial settings.

Multi-modal recommender system for predicting project manager performance within a competency-based framework

The evaluation of performance using competencies within a structured framework holds significant importance across various professional domains, particularly in roles like project manager. Typically, this assessment process, overseen by senior evaluators, involves scoring competencies based on data gathered from interviews, completed forms, and evaluation programs. However, this task is tedious and time-consuming, and requires the expertise of qualified professionals. Moreover, it is compounded by the inconsistent scoring biases introduced by different evaluators. In this paper, we propose a novel approach to automatically predict competency scores, thereby facilitating the assessment of project managers' performance. Initially, we performed data fusion to compile a comprehensive dataset from various sources and modalities, including demographic data, profile-related data, and historical competency assessments. Subsequently, NLP techniques were used to pre-process text data. Finally, recommender systems were explored to predict competency scores. We compared four different recommender system approaches: content-based filtering, demographic filtering, collaborative filtering, and hybrid filtering. Using assessment data collected from 38 project managers, encompassing scores across 67 different competencies, we evaluated the performance of each approach. Notably, the content-based approach yielded promising results, achieving a precision rate of 81.03%. Furthermore, we addressed the challenge of cold-starting, which in our context involves predicting scores for either a new project manager lacking competency data or a newly introduced competency without historical records. Our analysis revealed that demographic filtering achieved an average precision of 54.05% when dealing with new project managers. In contrast, content-based filtering exhibited remarkable performance, achieving a precision of 85.79% in predicting scores for new competencies. These findings underscore the potential of recommender systems in competency assessment, thereby facilitating more effective performance evaluation process.

An efficient deep learning model for tomato disease detection

Tomatoes possess significant nutritional and economic value. However, frequent diseases can detrimentally impact their quality and yield. Images of tomato diseases captured amidst intricate backgrounds are susceptible to environmental disturbances, presenting challenges in achieving precise detection and identification outcomes. This study focuses on tomato disease images within intricate settings, particularly emphasizing four prevalent diseases (late blight, gray leaf spot, brown rot, and leaf mold), alongside healthy tomatoes. It addresses challenges such as excessive interference, imprecise lesion localization for small targets, and heightened false-positive and false-negative rates in real-world tomato cultivation settings. To address these challenges, we introduce a novel method for tomato disease detection named TomatoDet. Initially, we devise a feature extraction module integrating Swin-DDETR’s self-attention mechanism to craft a backbone feature extraction network, enhancing the model’s capacity to capture details regarding small target diseases through self-attention. Subsequently, we incorporate the dynamic activation function Meta-ACON within the backbone network to further amplify the network’s ability to depict disease-related features. Finally, we propose an enhanced bidirectional weighted feature pyramid network (IBiFPN) for merging multi-scale features and feeding the feature maps extracted by the backbone network into the multi-scale feature fusion module. This enhancement elevates detection accuracy and effectively mitigates false positives and false negatives arising from overlapping and occluded disease targets within intricate backgrounds. Our approach demonstrates remarkable efficacy, achieving a mean Average Precision (mAP) of 92.3% on a curated dataset, marking an 8.7% point improvement over the baseline method. Additionally, it attains a detection speed of 46.6 frames per second (FPS), adeptly meeting the demands of agricultural scenarios.

Radar-optical fusion detection of UAV based on improved YOLOv7-tiny

This study presents a radar-optical fusion detection method for unmanned aerial vehicles (UAVs) in maritime environments. Radar and camera technologies are integrated to improve the detection capabilities of the platforms. The proposed method involves generating regions of interest (ROI) by projecting radar traces onto optical images through matrix transformation and geometric centroid registration. The generated ROI are matched with YOLO detection boxes using the intersection-over-union (IoU) algorithm, enabling radar-optical fusion detection. A modified algorithm, called SPN-YOLOv7-tiny, is developed to address the challenge of detecting small UAV targets that are easily missed in images. In this algorithm, the convolutional layers in the backbone network are replaced with a space-to-depth convolution, and a small object detection layer is added. In addition, the loss function was replaced with a normalized weighted distance loss function. Experimental results demonstrate that compared to the original YOLOv7-tiny method, SPN-YOLOv7-tiny achieves an improved mAP@0.5 (mean average precision at an IoU threshold of 0.5) from 0.852 to 0.93, while maintaining a high frame rate of 135.1 frames per second. Moreover, the proposed radar-optical fusion detection method achieves an accuracy of 96.98%, surpassing the individual detection results of the radar and camera. The proposed method effectively addresses the detection challenges posed by closely spaced overlapping targets on a radar chart.

A Comprehensive Evaluation of Dual-Polarimetric Sentinel-1 SAR Data for Monitoring Key Phenological Stages of Winter Wheat

Large-scale crop phenology monitoring is critical for agronomic planning and yield prediction applications. Synthetic Aperture Radar (SAR) remote sensing is well-suited for crop growth monitoring due to its nearly all-weather observation capability. Yet, the capability of the dual-polarimetric SAR data for wheat phenology estimation has not been thoroughly investigated. Here, we conducted a comprehensive evaluation of Sentinel-1 SAR polarimetric parameters’ sensibilities on winter wheat’s key phenophases while considering the incidence angle. We extracted 12 polarimetric parameters based on the covariance matrix and a dual-pol-version H-α decomposition. All parameters were evaluated by their temporal profile and feature importance score of Gini impurity with a decremental random forest classification process. A final wheat phenology classification model was built using the best indicator combination. The result shows that the Normalized Shannon Entropy (NSE), Degree of Linear Polarization (DoLP), and Stokes Parameter g2 were the three most important indicators, while the Span, Average Alpha (α2¯), and Backscatter Coefficient σVH0 were the three least important features in discriminating wheat phenology for all three incidence angle groups. The smaller-incidence angle (30–35°) SAR images are better suited for estimating wheat phenology. The combination of NSE, DoLP, and two Stokes Parameters (g2 and g0) constitutes the most effective indicator ensemble. For all eight key phenophases, the average Precision and Recall scores were above 0.8. This study highlighted the potential of dual-polarimetric SAR data for wheat phenology estimation. The feature importance evaluation results provide a reference for future phenology estimation studies using dual-polarimetric SAR data in choosing better-informed indicators.

Signal alignment for cross-datasets in P300 brain-computer interfaces

Objective. Transfer learning has become an important issue in the brain-computer interface (BCI) field, and studies on subject-to-subject transfer within the same dataset have been performed. However, few studies have been performed on dataset-to-dataset transfer, including paradigm-to-paradigm transfer. In this study, we propose a signal alignment (SA) for P300 event-related potential (ERP) signals that is intuitive, simple, computationally less expensive, and can be used for cross-dataset transfer learning. Approach. We proposed a linear SA that uses the P300’s latency, amplitude scale, and reverse factor to transform signals. For evaluation, four datasets were introduced (two from conventional P300 Speller BCIs, one from a P300 Speller with face stimuli, and the last from a standard auditory oddball paradigm). Results. Although the standard approach without SA had an average precision (AP) score of 25.5%, the approach demonstrated a 35.8% AP score, and we observed that the number of subjects showing improvement was 36.0% on average. Particularly, we confirmed that the Speller dataset with face stimuli was more comparable with other datasets. Significance. We proposed a simple and intuitive way to align ERP signals that uses the characteristics of ERP signals. The results demonstrated the feasibility of cross-dataset transfer learning even between datasets with different paradigms.

A New Dataset and Comparative Study for Aphid Cluster Detection and Segmentation in Sorghum Fields

Aphid infestations are one of the primary causes of extensive damage to wheat and sorghum fields and are one of the most common vectors for plant viruses, resulting in significant agricultural yield losses. To address this problem, farmers often employ the inefficient use of harmful chemical pesticides that have negative health and environmental impacts. As a result, a large amount of pesticide is wasted on areas without significant pest infestation. This brings to attention the urgent need for an intelligent autonomous system that can locate and spray sufficiently large infestations selectively within the complex crop canopies. We have developed a large multi-scale dataset for aphid cluster detection and segmentation, collected from actual sorghum fields and meticulously annotated to include clusters of aphids. Our dataset comprises a total of 54,742 image patches, showcasing a variety of viewpoints, diverse lighting conditions, and multiple scales, highlighting its effectiveness for real-world applications. In this study, we trained and evaluated four real-time semantic segmentation models and three object detection models specifically for aphid cluster segmentation and detection. Considering the balance between accuracy and efficiency, Fast-SCNN delivered the most effective segmentation results, achieving 80.46% mean precision, 81.21% mean recall, and 91.66 frames per second (FPS). For object detection, RT-DETR exhibited the best overall performance with a 61.63% mean average precision (mAP), 92.6% mean recall, and 72.55 on an NVIDIA V100 GPU. Our experiments further indicate that aphid cluster segmentation is more suitable for assessing aphid infestations than using detection models.

Automated Detection and Counting of Wild Boar in Camera Trap Images

Camera traps are becoming widely used for wildlife monitoring and management. However, manual analysis of the resulting image sets is labor-intensive, time-consuming and costly. This study shows that automated computer vision techniques can be extremely helpful in this regard, as they can rapidly and automatically extract valuable information from the images. Specific training with a set of 1600 images obtained from a study where wild animals approaching wild boar carcasses were monitored enabled the model to detect five different classes of animals automatically in their natural environment with a mean average precision of 98.11%, namely ‘wild boar’, ‘fox’, ‘raccoon dog’, ‘deer’ and ‘bird’. In addition, sequences of images were automatically analyzed and the number of wild boar visits and respective group sizes were determined. This study may help to improve and speed up the monitoring of the potential spread of African swine fever virus in areas where wild boar are affected.

A bearing surface defect detection method based on multi-attention mechanism Yolov8

Surface defects in bearings not only affect the appearance but also impact the service life and performance. Therefore, it is imperative for bearing manufacturers to conduct quality inspections before bearings leave the factory. However, traditional visual inspection methods exhibit shortcomings such as high omission rates, insufficient feature fusion and oversized models when dealing with multiple target defects in bearings. To address these challenges, this paper proposes a surface defect detection method for bearings based on an improved Yolov8 algorithm (G-Yolov8). Firstly, a C3Ghost convolutional module based on the Ghost module is constructed in YOLOv8 to simplify model computational costs. Secondly, a global attention mechanism module is designed at the end of the backbone network to increase sensitivity to implicit small target area features and optimize feature extraction efficiency. Subsequently, a deep deformable convolution feature pyramid network is constructed by introducing the deformable convolutional networks version 2 (DCNv2) and the lightweight content-aware reassembly of features upsampling operator to reduce sampling information loss and improve the fusion of multi-scale target defects. Finally, different attention mechanisms are embedded in the detection network to construct a multi-attention detection head to replace the decoupled head, refining classification and localization tasks, reducing feature confusion, and improving the model’s detection accuracy. Experimental results demonstrate that the improved algorithm achieves a 3.5% increase in mean average precision on a self-made small-scale train bearing surface defect dataset, with a 17.3% reduction in model size. This improvement not only enhances accuracy but also addresses the requirement for lightweight deployment in subsequent stages.

Real-Time Detection Technology of Corn Kernel Breakage and Mildew Based on Improved YOLOv5s

In order to solve low recognition of corn kernel breakage degree and corn kernel mildew degree during corn kernel harvesting, this paper proposes a real-time detection method for corn kernel breakage and mildew based on improved YOlOv5s, which is referred to as the CST-YOLOv5s model algorithm in this paper. The method continuously obtains images through the discrete uniform sampling device of corn kernels and generates whole corn kernels, breakage corn kernels, and mildew corn kernel dataset samples. We aimed at the problems of high similarity of some corn kernel features in the acquired images and the low precision of corn kernel breakage and mildew recognition. Firstly, the CBAM attention mechanism is added to the backbone network of YOLOv5s to finely allocate and process the feature information, highlighting the features of corn breakage and mildew. Secondly, the pyramid pooling structure SPPCPSC, which integrates cross-stage local networks, is adopted to replace the SPPF in YOLOv5s. SPP and CPSC technologies are used to extract and fuse features of different scales, improving the precision of object detection. Finally, the original prediction head is converted into a transformer prediction head to explore the prediction potential with a multi-head attention mechanism. The experimental results show that the CST-YOLOv5s model has a significant improvement in the detection of corn kernel breakage and mildew. Compared with the original YOLOv5s model, the average precision (AP) of corn kernel breakage and mildew recognition increased by 5.2% and 7.1%, respectively, and the mean average precision (mAP) of all kinds of corn kernel recognition is 96.1%, and the frame rate is 36.7 FPS. Compared with YOLOv4-tiny, YOLOv6n, YOLOv7, YOLOv8s, and YOLOv9-E detection model algorithms, the CST-YOLOv5s model has better overall performance in terms of detection accuracy and speed. This study can provide a reference for real-time detection of breakage and mildew kernels during the harvesting process of corn kernels.

CoMadOut—a robust outlier detection algorithm based on CoMAD

AbstractUnsupervised learning methods are well established in the area of anomaly detection and achieve state of the art performances on outlier datasets. Outliers play a significant role, since they bear the potential to distort the predictions of a machine learning algorithm on a given dataset. Especially among PCA-based methods, outliers have an additional destructive potential regarding the result: they may not only distort the orientation and translation of the principal components, they also make it more complicated to detect outliers. To address this problem, we propose the robust outlier detection algorithm CoMadOut, which satisfies two required properties: (1) being robust towards outliers and (2) detecting them. Our CoMadOut outlier detection variants using comedian PCA define, dependent on its variant, an inlier region with a robust noise margin by measures of in-distribution (variant CMO) and optimized scores by measures of out-of-distribution (variants CMO*), e.g. kurtosis-weighting by CMO+k. These measures allow distribution based outlier scoring for each principal component, and thus, an appropriate alignment of the degree of outlierness between normal and abnormal instances. Experiments comparing CoMadOut with traditional, deep and other comparable robust outlier detection methods showed that the performance of the introduced CoMadOut approach is competitive to well established methods related to average precision (AP), area under the precision recall curve (AUPRC) and area under the receiver operating characteristic (AUROC) curve. In summary our approach can be seen as a robust alternative for outlier detection tasks.

Computer vision digitization of smartphone images of anesthesia paper health records from low-middle income countries

BackgroundIn low-middle income countries, healthcare providers primarily use paper health records for capturing data. Paper health records are utilized predominately due to the prohibitive cost of acquisition and maintenance of automated data capture devices and electronic medical records. Data recorded on paper health records is not easily accessible in a digital format to healthcare providers. The lack of real time accessible digital data limits healthcare providers, researchers, and quality improvement champions to leverage data to improve patient outcomes. In this project, we demonstrate the novel use of computer vision software to digitize handwritten intraoperative data elements from smartphone photographs of paper anesthesia charts from the University Teaching Hospital of Kigali. We specifically report our approach to digitize checkbox data, symbol-denoted systolic and diastolic blood pressure, and physiological data.MethodsWe implemented approaches for removing perspective distortions from smartphone photographs, removing shadows, and improving image readability through morphological operations. YOLOv8 models were used to deconstruct the anesthesia paper chart into specific data sections. Handwritten blood pressure symbols and physiological data were identified, and values were assigned using deep neural networks. Our work builds upon the contributions of previous research by improving upon their methods, updating the deep learning models to newer architectures, as well as consolidating them into a single piece of software.ResultsThe model for extracting the sections of the anesthesia paper chart achieved an average box precision of 0.99, an average box recall of 0.99, and an mAP0.5-95 of 0.97. Our software digitizes checkbox data with greater than 99% accuracy and digitizes blood pressure data with a mean average error of 1.0 and 1.36 mmHg for systolic and diastolic blood pressure respectively. Overall accuracy for physiological data which includes oxygen saturation, inspired oxygen concentration and end tidal carbon dioxide concentration was 85.2%.ConclusionsWe demonstrate that under normal photography conditions we can digitize checkbox, blood pressure and physiological data to within human accuracy when provided legible handwriting. Our contributions provide improved access to digital data to healthcare practitioners in low-middle income countries.

A convolutional neural network (Seg-CNN) for recognition, classifi cation, and segmentation of apple fruit branches and stems in images

This article reports the results of research studies conducted in 2023–2024 on transfer learning of Segmentation Convolutional Neural Networks (Seg-CNN) models for classification, recognition, and segmentation of branches with apple fruits and stems in images. State-of-the-art convolutional neural network architectures, i.e., YOLOv8(n,s,m,l,x)-seg, were used for a detailed segmentation of biological objects in images of varying complexity and scale at the pixel level. An image dataset collected in the fi eld using a GoPro HERO 11 camera was marked up for transfer model training. Data augmentation was performed, producing a total of 2500 images. Image markup was performed using the polygon annotation tool. As a result, polygonal contours around objects were created, outlines of branches, apple tree fruits, and stems were outlined, and segments of objects in the images were indicated. The objects were assigned the following classes: Apple branch, Apple fruit, and Apple stem. Binary classification metrics, such as Precision and Recall, as well as Mean Average Precision (mAP), were used to evaluate the performance of the trained models in recognizing branches with apple fruits and stems in images. The YOLOv8x-seg (mAP50 0.758) and YOLOv8l-seg (mAP50 0.74) models showed high performance in terms of all metrics in recognizing branches, apple fruit, and fruit stems in images, outperforming the YOLOv8n-seg (mAP50 0.7) model due to their more complex architecture. The YOLOv8n-seg model has a faster frame processing speed (11.39 frames/s), rendering it a preferred choice for computing systems with limited resources. The results obtained confirm the prospects of using machine learning algorithms and convolutional neural networks for segmentation and pixel-by-pixel classification of branches with apple fruits and stems on RGB images for monitoring the condition of plants and determining their geometric characteristics.

Effective small crack detection based on tunnel crack characteristics and an anchor-free convolutional neural network

Tunnel cracks are thin and narrow linear targets, and their pixel proportions in images are usually very low, less than 6%; therefore, a method is needed to better detect small crack targets. In this study, a crack detection method based on crack characteristics and an anchor-free framework is investigated. First, the characteristics of cracks are analyzed to obtain the real crack texture, interference noise texture, and targets appearing near each crack as the context information for the model to filter and remove noise. We discuss the crack detection performance of anchor-based and anchor-free algorithms. Then, an optimized anchor-free algorithm is proposed in this paper for crack detection. Based on the advantages of YOLOX-x, we add a semantic enhancement module to better use contextual information. The experimental results show that the anchor-free algorithm performs slightly better than other algorithms in crack detection situations. In addition, the proposed method displays better detection performance for slender and inconspicuous cracks, with an average precision of 0.858.

Deep learning-based bubble detection with swin transformer

ABSTRACT We developed a deep learning-based bubble detector with a Shifted window Transformer (Swin Transformer) to detect and segment individual bubbles among overlapping bubbles. To verify the performance of the detector, we calculated its average precision (AP) with different number of training images. The mask AP increased with the increase in the number of training images when there were less than 50 images. It was observed that the AP for the Swin Transformer and ResNet were almost the same when there were more than 50 images; however, when few training images were used, the AP of the Swin Transformer were higher than that of the ResNet. Furthermore, for the increase in void fraction, the AP of the Swin Transformer showed a decrease similar to that in the case of the ResNet; however, for few training images, the AP of the Swin Transformer was higher than that of the ResNet in all void fractions. Moreover, we confirmed the detector trained with experimental and synthetic bubble images was able to segment overlapping bubbles and deformed bubbles in a bubbly flow experiment. Thus, we verified that the new bubble detector with Swin Transformer provided higher AP than the detector with ResNet for fewer training images.