Matching Accuracy Research Articles

Monitoring Use of Language Interpreting Services for Patients with Limited English Proficiency: Methods to Match Patient Medical Records with Interpreter Billing Logs.

Over 60 million patients in the USA have limited English proficiency (LEP) and experience barriers in care. Still, there exists no standardized method of monitoring the utilization of language interpreting services (LIS). To introduce a methodological approach to systematically monitor utilization of LIS for LEP patients. We utilized a One-To-Many Match algorithm to align inpatient visits of LEP patients from the electronic health record (EHR) with corresponding calls from LIS billing logs, using a unique patient identifier (MRN) and LIS call dates within patient's admit and discharge dates. Due to error when MRNs are recorded by LIS, the FuzzyWuzzy Probabilistic String-Matching technique was utilized to enhance match accuracy where exact matches were unattainable, addressing inherent complexities in language data matching. The study involved 5823 inpatient encounters with a non-English preference in an urban hospital system in 2020, representing a linguistically diverse patient base, and attempted to match these against 183,655 LIS call logs. Our approach successfully matched 83.1% (4389 out of 5823) of inpatient encounters to an LIS call. We observed significant language-specific disparities in LIS usage, with Spanish leading in call volume at 2737 calls (exact matches) and 845 (probabilistic matches). Concordance rates varied, exceeding 94% for all languages in exact matches and ranging from 53.9% for Arabic to 71.6% for Russian in probabilistic matches. The average frequency of LIS calls was about one call per day per language group in the inpatient setting. The study provides vital insights into language service preferences, frequency, and duration.These findings emphasize the need for standard methods in monitoring LIS usage to enhance patient outcomes for LEP patients.

Establishment and extension of a compact and robust binary feature descriptor for UAV image matching

Abstract Local feature description is a crucial challenge in unmanned aerial vehicle (UAV) images matching. Although scholars have explored a variety of methods for defining local feature descriptions, improving the matching accuracy of UAV images and reducing the memory consumption of descriptor remain worthwhile research issues. A compact and robust binary feature descriptor for UAV image matching is proposed. To achieve rotational invariance of the descriptor, a sampling pattern based on the region overlap error is employed to assign the dominant direction of the feature points. Additionally, a rotation invariant gradient computation method is proposed to reduce the dominant direction assignment errors. Then, the region around the feature point is divided into grids based on the region overlap error, and gradient orientation histogram and average intensity histogram are constructed. Finally, these histograms are binarized to generate a 30-byte binary descriptor. Results from the Oxford dataset, the Iguazu dataset, and a high-resolution UAV image dataset indicate that the proposed descriptor is compact and achieves the highest matching accuracy compared to nine commonly used feature descriptors, including scale-invariant feature transform, speeded up robust feature, binary robust invariant scalable keypoints, KAZE, oriented fast and rotated brief, fast retina keypoint, local intensity order pattern, boosted efficient binary local image descriptor, and triplet-based efficient binary local image descriptor. Notably, on the high-resolution UAV images with translational transformations, scale variations, salt-and-pepper noise, and Gaussian white noise, the accuracies of the proposed descriptor were 87.09%, 80.83%, 92.46%, and 87.42%, respectively. The improvements ranged from 6.95% to 53.20%, 9.32% to 55.90%, 10.39% to 44.29%, and 6.54% to 58.24%.

Application Introduction and Learning Geometry of Space to Student Elementary School Using Method Rolling Hash Based on Android

Along with technological advancements, smartphones have evolved from being just a means of communication to now being multifunctional devices that reflect the status of their users and offer various features, such as cameras, music and video players, video calls, voice recorders, games, and internet access. These innovations' progress made them widely used by humans in their activities. One of the innovations that utilizes this technology is a building space introduction and learning application explicitly designed for elementary school students, using the Android-based Rolling Hash method. This application is an interactive educational tool that provides comprehensive information about building spaces, explains the various forms of building spaces, provides applicable examples, and offers solutions to related problems. This application's primary focus is implementing the Rolling Hash algorithm for string matching, which was chosen for its ability to perform fast and accurate matching in answering questions. By implementing this algorithm, the application can efficiently handle various questions involving strings, enhancing students' learning experience. This application uses Java programming language with ADT IDE for Eclipse and SQLite as a database system to store and manage data. The result of this final project is a learning application that not only facilitates the introduction of building space concepts to elementary school students but also shows how modern technology can improve learning methods. The Rolling Hash algorithm proved its effectiveness in fast and precise string matching, providing an optimal solution for this educational application and presenting a model of how technological tools can support primary education innovatively and practically.

The application of artificial intelligence in biomechanical feedback and learning effectiveness enhancement in ideological and political education

The application of artificial intelligence (AI) in ideological and political education (IPE) optimizes the learning effect with the help of biomechanical feedback mechanism, and provides technical support for deepening the reform of IPE in colleges and universities. With the help of AI technology, especially its advantages in data analysis and deep learning (DL), accurate matching and personalized recommendation of learning resources can be realized. In this study, an AI-assisted IPE system with biomechanical feedback was constructed. By analyzing students’ physiological reactions and behavior patterns in the learning process, the system can evaluate the learning state and adjust teaching strategies. The key findings show that this learning method combined with biomechanical feedback can significantly improve students’ learning participation, understanding depth and satisfaction, and make IPE more in line with students’ needs.

Cross-Domain Person Re-Identification Based on Multi-Branch Pose-Guided Occlusion Generation.

Aiming at the problems caused by a lack of feature matching due to occlusion and fixed model parameters in cross-domain person re-identification, a method based on multi-branch pose-guided occlusion generation is proposed. This method can effectively improve the accuracy of person matching and enable identity matching even when pedestrian features are misaligned. Firstly, a novel pose-guided occlusion generation module is designed to enhance the model's ability to extract discriminative features from non-occluded areas. Occlusion data are generated to simulate occluded person images. This improves the model's learning ability and addresses the issue of misidentifying occlusion samples. Secondly, a multi-branch feature fusion structure is constructed. By fusing different feature information from the global and occlusion branches, the diversity of features is enriched. This enrichment improves the model's generalization. Finally, a dynamic convolution kernel is constructed to calculate the similarity between images. This approach achieves effective point-to-point matching and resolves the problem of fixed model parameters. Experimental results indicate that, compared to current mainstream algorithms, this method shows significant advantages in the first hit rate (Rank-1), mean average precision (mAP), and generalization performance. In the MSMT17→DukeMTMC-reID dataset, after re-ranking (Rerank) and time-tift (Tlift) for the two indicators on Market1501, the mAP and Rank-1 reached 80.5%, 84.3%, 81.9%, and 93.1%. Additionally, the algorithm achieved 51.6% and 41.3% on DukeMTMC-reID→Occluded-Duke, demonstrating good recognition performance on the occlusion dataset.

A clip-based dual-stream method for text based vehicle search

A text-based vehicle search refers to a system where users can find vehicles or route information by entering text-based queries. The primary objective of text-based vehicle search is to identify the most relevant vehicle in a given dataset using a natural language description as a query. This approach leverages natural language processing (NLP) to understand and interpret description queries and provide relevant results. Despite significant progress, this task still faces several challenges due to the complexity and diversity of natural language, as well as inherent difficulties in the vision domain. Moreover, few studies have focused on tracked-vehicle retrieval, where vehicle tracklets are considered instead of single images. In this paper, we propose a novel framework for natural language-based tracked-vehicle retrieval based on CLIP model, one of the most effective models for image-text matching task. This framework leverages both appearance and motion information to enhance the matching accuracy of vehicle tracklet retrieval. Some experiments are conducted on the CityFlow-NL dataset, provided by the 6-th AI City Challenge, an annual competition. The results are comparable to state-of-the-art methods, achieving an MRR score of 46.63%, Rank@5 of 67.02%, and Rank@10 of 81.82%

An Infrared and Visible Image Alignment Method Based on Gradient Distribution Properties and Scale-Invariant Features in Electric Power Scenes.

In grid intelligent inspection systems, automatic registration of infrared and visible light images in power scenes is a crucial research technology. Since there are obvious differences in key attributes between visible and infrared images, direct alignment is often difficult to achieve the expected results. To overcome the high difficulty of aligning infrared and visible light images, an image alignment method is proposed in this paper. First, we use the Sobel operator to extract the edge information of the image pair. Second, the feature points in the edges are recognised by a curvature scale space (CSS) corner detector. Third, the Histogram of Orientation Gradients (HOG) is extracted as the gradient distribution characteristics of the feature points, which are normalised with the Scale Invariant Feature Transform (SIFT) algorithm to form feature descriptors. Finally, initial matching and accurate matching are achieved by the improved fast approximate nearest-neighbour matching method and adaptive thresholding, respectively. Experiments show that this method can robustly match the feature points of image pairs under rotation, scale, and viewpoint differences, and achieves excellent matching results.

Speech Production of Mandarin Lexical Tones Among Canadian Elementary Students Enrolled in Mandarin-English Bilingual Schools.

This study investigates how Mandarin-English bilingual students in Canada produce Mandarin tones and how this is influenced by factors such as tone complexity, cross-linguistic influences, and speech input. Participants were 82 students enrolled in a Chinese bilingual program in Western Canada. Students were recruited from Grades 1, 3, and 5 and divided into two groups based on their home language backgrounds: The heritage language group had early and strong input in Mandarin, and the second language (L2) group received mostly English input at home. Single-word tone productions were audio-recorded and transcribed by Mandarin-native listeners for match (accuracy) and pattern analyses. Acoustic measurements were extracted to provide phonetic details. First, Tone3 (dipping tone) was challenging across groups due to its complexity. Second, L2 students' productions were more influenced by English as a nontonal language and showed signs of categorical confusion. Third, increased tone match rates were related to both home input and school input, but bilingual students did not reach more than 90% of match rates in Grade 5. Instead, L2 students produced phonetic features less accurately in higher grades. This was attributed to reduced pronunciation instruction and limited home input. Bilingual students' speech development in a minority language indicates unique influences of home and school input but also the universal influences of tone complexity. This study provides evidence for bilingual speech theories in the suprasegmental domain and has implications for the pedagogy of a minority language in the context of bilingual education. https://doi.org/10.23641/asha.28098206.

POI Data Fusion Method Based on Multi-Feature Matching and Optimization

The key to geospatial data integration lies in identifying corresponding objects from different sources. Aiming at the problem of the low matching accuracy of geospatial entities under a single feature attribute, a geospatial entity matching method based on multi-feature value calculation is proposed. Firstly, when dealing with POI (point of interest) data, the similarity of POI data in terms of name, address, and distance is calculated by combining the improved hybrid similarity method, the Jaccard method, and the Euclidean metric method. Secondly, the random forest algorithm is utilized to dynamically determine the information weights of each attribute and calculate the comprehensive similarity. Finally, taking the area within the Second Ring Road in Beijing as the experimental area, the POI data of Tencent Maps and Amap are collected to verify the method proposed in this paper. The experimental results show that, compared with the existing POI matching methods, the accuracy and recall rate of the results obtained by the POI matching and fusion method proposed in this paper are significantly improved, which verifies the accuracy and feasibility of the matching.

Plucking Point and Posture Determination of Tea Buds Based on Deep Learning

Tea is a significant cash crop grown widely around the world. Currently, tea plucking predominantly relies on manual work. However, due to the aging population and increasing labor costs, machine plucking has become an important trend in the tea industry. The determination of the plucking position and plucking posture is a critical prerequisite for machine plucking tea leaves. In order to improve the accuracy and efficiency of machine plucking tea leaves, a method is presented in this paper to determine the plucking point and plucking posture based on the instance segmentation deep learning network. In this study, tea images in the dataset were first labeled using the Labelme software (version 4.5.13), and then the LDS-YOLOv8-seg model was proposed to identify the tea bud region and plucking area. The plucking points and the central points of the tea bud’s bounding box were calculated and matched as pairs using the nearest point method (NPM) and the point in range method (PIRM) proposed in this study. Finally, the plucking posture was obtained according to the results of the feature points matching. The matching results on the test dataset show that the PIRM has superior performance, with a matching accuracy of 99.229% and an average matching time of 2.363 milliseconds. In addition, failure cases of feature points matching in the plucking posture determination process were also analyzed in this study. The test results show that the plucking position and posture determination method proposed in this paper is feasible for machine plucking tea.

Job matching analysis by latent semantic indexing enhanced on multilingual word meanings

Job matching is a hiring process that involves a thorough understanding of the context and meaning of words in different languages. The updated and expanded latent semantic indexing (LSI) Framework seeks to improve the precision and relevance of job matching analysis of word meanings in multi-languages. Because they only compare related terms, conventional LSIs are often insufficient to address the complexity of context in job matching. Extending the LSI approach can improve the vector representation of words and help you understand the context and semantic relationships in the text. Improved LSI analyzes context more precisely by using word vector representation. Improved LSI focuses on understanding semantic relationships between words in many languages to produce more accurate and relevant job matches. This paper describes the steps involved in improving LSI, such as data collection, pre-processing, linguistic feature extraction, LSI model training, and evaluation of matching results. The results show that the examined classification model has much better performance in terms of word classification. Conventional LSI has an average prediction value of 79%, once the enhanced LSI can accurately predict about 84% of the entire word, it has a reasonable capacity to recognize the actual words in a natural context.

Boarding stop inference with uncertain relationship between bus vehicles and mobile smart card readers

AbstractBoarding stop inference for bus passengers is essential for the improvement of bus transit services. Previous studies mainly focus on matching the bus trajectories with the bus stop locations, while the relationship between smart card readers—which collect the smart card data—and bus vehicles is usually given. However, uncertainties arise in practical applications regarding the matching of vehicles and card readers. To tackle this challenge, in this study, a data‐driven approach is proposed to dig into the spatiotemporal features of passengers' smart card data and bus vehicle operations. A weighted bipartite graph algorithm is developed to match the smart card readers with the bus vehicles automatically. To verify the feasibility and effectiveness of the proposed approach, a case study is conducted on the Bus Anhong Line in Shanghai, China. The inferred results of boarding stops are compared with the data from passenger counting sensors installed in the bus vehicles. The matching accuracy rate achieves 0.9539, which validates the effectiveness of the proposed matching model. In addition, the inferred data are used to present the spatiotemporal patterns of boarding passengers and identify high‐demand bus stops.

A Novel Map Matching Method Based on Improved Hidden Markov and Conditional Random Fields Model

ABSTRACT Aiming at the inherent “labeling bias” problem of Hidden Markov Models (HMMs) in the process of low-frequency sampling rate GPS trajectory map matching, this paper proposes a map matching method that combines HMMs and Conditional Random Field Models (CRFs). Various features including vehicle driving direction, direction change between candidate roads, shortest path, and spatial localization accuracy of trajectory points are integrated to optimize the state transfer process in the HMM model. The Viterbi algorithm was then employed to calculate the joint probability values of all candidate paths and the path with maximum probability value was selected as the matching path. The proposed map matching method was validated using publicly available Global Positioning System (GPS) vehicle trajectories from Guangzhou, Dongguan, and field-collected vehicle trajectory data from Ganzhou, China. The results demonstrated that the proposed method achieves high matching accuracy while maintaining efficient matching efficiency. The percentage of correct matching of GPS sampling points remains above 95%, and the average calculation time for each candidate road segment is kept within 50 ms. These advancements can greatly benefit location-based services that rely on map matching technology.

Design Framework of Expert System Program in Otolaryngology Disease Diagnosis use Extreme Programming (XP)Method(Case Study in THB Bekasi Hospital)

The prevalence of Ear, Nose, and Throat (ENT) diseases presents diagnostic challenges, especially in resource-limited settings. At THB Bekasi Hospital, constrained specialist availability and long consultation queues highlight the need for an accessible diagnostic solution. This study aims to develop an expert system for diagnosing ENT diseases using the Extreme Programming (XP) methodology, incorporating the forward chaining technique for inference. The research includes assessment, knowledge acquisition through expert consultations, system design, and rigorous testing. The system was developed as a mobile application using Android Studio, enabling users to input symptoms and receive real-time diagnostic insights. The knowledge base integrates data from medical experts, synthesizing 11 diseases and 35 symptoms into a robust decision-making framework. The diagnostic process applies predefined rules to ensure accuracy in identifying conditions such as sinusitis, laryngitis, and otitis. Evaluation results demonstrate a 100% match accuracy during testing with 15 test cases, confirming the system’s reliability. The application offers users rapid diagnostic assistance, promoting timely treatment for ENT issues, although it does not substitute medical professionals. By leveraging ubiquitous smartphone access, this system addresses gaps in healthcare accessibility and enhances patient autonomy. This research contributes a scalable framework for deploying expert systems in other medical domains. Future improvements include integrating geolocation services for nearby specialist referrals and adopting backward chaining for more complex diagnoses, thereby broadening its applicability and utility

IMAGE MATCHING BASED HAZARDOUS MATERIAL DETECTION AND WARNING SYSTEM

Transportation of dangerous goods involves many critical situations that require safety and special precautions. In accordance with the regulations, hazardous materials, which include international standards, should be closely monitored and precautions should be taken in advance according to the situation. Artificial intelligence, image processing and data analysis techniques can be used to recognize and classify the labels of dangerous goods. This is important for early action in case of an emergency. If hazardous materials are not properly stored or transported according to safety precautions and rules, they can cause both material and moral damage. In this study, a hazardous material detection and warning system using AKAZE, ORB and SIFT image feature matching techniques is developed. To test the system, a dataset of multiple hazardous material labels with different scenes and conditions was created. The performances of feature matching techniques including image processing algorithms are examined through comparative analysis. As a result of image matching, label-related features and intervention information were retrieved from the database and displayed on the system interface. Experimental results show that the ORB technique is the best method for feature matching and accurate matching, and the AKAZE technique is the fastest feature detection method.

Dual-Branch Feature Generalization Method for AUV Near-Field Exploration of Hydrothermal Areas

The simultaneous localization and mapping (SLAM) technique provides long-term near-seafloor navigation for autonomous underwater vehicles (AUVs). However, the stability of descriptors generated by interest point detectors remains a challenge in the hydrothermal environment. This paper proposes a dual-branch feature generalization method, incorporating volumetric density and color distribution for enhanced robustness. The method utilizes shared descriptors and a feature confidence mechanism, combining neural radiance fields with Gaussian splatting models, ensuring fast and accurate feature generalization. The proposed approach improves recall while maintaining matching accuracy, ensuring stability and robustness in feature matching. This method achieves stable and reliable feature matching in a simulated hydrothermal environment.

Map-Matching Algorithm Based on Hidden Markov and Constraint Value Pruning

Map matching is the process of matching global positioning system (GPS) trajectory data with map data. Its purpose is to determine the actual route of the moving object. Because of factors such as positioning devices and the environment, the GPS trajectory data obtained may not be accurate. In location-based services, map matching can be used to address the accuracy and reliability issues of GPS location data. There are currently many map-matching algorithms, for example, spatio-temporal-based (STD) matching algorithm, improved interactive voting-based map-matching (IIVMM) algorithm, and turning-point-based (TPB) offline map-matching algorithm, but existing algorithms have some shortcomings, such as low matching accuracy on complex roads or low sampling rates, and routing calculation time bottlenecks. Therefore, this paper proposes a fast matching algorithm based on constrained value pruning that is suitable for complex roads. The algorithm considers the multi-directional features within the road and uses secondary calculations to determine candidate points, improving the accuracy of candidate point selection. Additionally, two pruning strategies based on constrained values are introduced to reduce the majority of the routing calculation process and improve the matching efficiency. Finally, comparative experiments are conducted on a real trajectory dataset. The results show that, compared with STD, IIVMM, and TPB algorithms, the algorithm accuracy is improved by about 2% to 4%, and the running time is reduced by about 30%.

Effects of attentional focus on spatial localization of distal body parts and touch in two-arm position matching.

This study investigated how the judgment of proximal joint position can be affected by touch alone, focused attention on the distal body part, or touch spatial localization. Participants completed a two-arm elbow joint position-matching task, in which they indicated the location of one forearm by the placement of the other. In four test conditions, matching was performed during (1) detection of touch (tactile stimulation of index finger pads), (2) spatial localization of fingers (attention focused on the position of index finger pads), (3) spatial localization of touch on fingers (attention focused on tactile stimulation of index finger pads), and (4) detection of touch but localization of fingers (tactile stimulation of index finger pads, but attention focusing on the spatial position of the pads). In the first experiment (n = 23), the sensitivity of muscle spindle receptors in both reference and indicator arms was reduced and equalized by both-slack conditioning. In the second experiment (n = 20), the illusion of excessive elbow flexion in the reference arm and excessive extension in the indicator arm was generated through extension-flexion conditioning. In the first experiment, the accuracy and precision of matching were unaffected in any test condition. In the second experiment, participants made amplified undershooting errors under attention-focused conditions. In conclusion, focused attention on the location of a distal body part and touch affects both the spatial localization of the limb and tactile remapping only when the perceived forearm position is misinterpreted due to imbalanced proprioceptive input from antagonistic arm muscles.

Face and voice identity matching accuracy is not improved by multimodal identity information.

Identity verification from both faces and voices can be error-prone. Previous research has shown that faces and voices signal concordant information and cross-modal unfamiliar face-to-voice matching is possible, albeit often with low accuracy. In the current study, we ask whether performance on a face or voice identity matching task can be improved by using multimodal stimuli which add a second modality (voice or face). We find that overall accuracy is higher for face matching than for voice matching. However, contrary to predictions, presenting one unimodal and one multimodal stimulus within a matching task did not improve face or voice matching compared to presenting two unimodal stimuli. Additionally, we find that presenting two multimodal stimuli does not improve accuracy compared to presenting two unimodal face stimuli. Thus, multimodal information does not improve accuracy. However, intriguingly, we find that cross-modal face-voice matching accuracy predicts voice matching accuracy but not face matching accuracy. This suggests cross-modal information can nonetheless play a role in identity matching, and face and voice information combine to inform matching decisions. We discuss our findings in light of current models of person perception, and consider the implications for identity verification in security and forensic settings.

Detection of Helmet Violations among Electric Bicycle Riders Through Multi-Network

The use of helmets plays a crucial role in mitigating head injuries resulting from traffic accidents. However, manual supervision of helmet usage is both costly and prone to missing target objects. With the advancement of technology and widespread use of surveillance cameras, it has become technically feasible to automatically detect riders’ helmet wearing based on roadside videos. This capability holds significant potential in effectively guiding riders to wear helmets and enhancing driving safety. This paper presents a multi-network-based approach for automated detection of helmet wearing among electric bicycle riders. The object detection model is employed to detect electric bicycles and helmets. Additionally, the overall approach utilizes a human pose estimator to detect keypoints of human skeletons for selecting the region of interest (RoI) corresponding to the head region location. We propose location rules to extract the rider region within the electric bicycle box for distinguishing riders and locate the head region based on human skeleton keypoints for helmet wearing matching. The mean average precision (mAP) of the detection model trained on public datasets reaches 91.8%. To evaluate matching accuracy, we randomly selected 150 correctly detected images from the open-sourced two-wheeler helmet dataset (TWHD), achieving a matching accuracy of 95.6%. The detection model is trained and tested on public datasets which effectively facilitates the processes of data collection and annotation. Based on multi-network information fusion, we can accurately detect helmet violations among electric bicycle riders in multi-person interaction scenarios.