Improve Identification Accuracy Research Articles

Characteristics and Source Analysis of Ozone Pollution in Tianjin from 2013 to 2022

This study has analyzed ozone pollution in Tianjin from 2013 to 2022, focusing on the relationships between ozone distribution, meteorological conditions, and precursor substances. A method for identifying high-value areas of ozone precursors using the Ozone Sensitivity Factor (FNR) has been introduced. Results show that the average ozone concentration in Tianjin has been 100.608 µg/m3, with an annual growth rate of 2.84 µg·m⁻3·yr⁻¹. Tianjin has ranked among the top provinces and urban agglomerations in China for both ozone concentration and growth rate. Ozone levels have peaked in summer, followed by spring, autumn, and winter, while the growth rate has been highest in spring. This indicates that ozone pollution extends from summer into spring and autumn. An analysis of six ozone pollution events reveals significant regional transmission impacts from northern Hebei and Inner Mongolia, contributing over 30%, with additional significant contributions from southern and southwestern Hebei and western Shandong. In terms of controlling ozone precursors, high-HCHO-value areas have been identified. The correlation between areas of high HCHO values and ground-level ozone concentrations was 0.56339 during the ozone season and 0.2214 during the non-ozone season, both of which improved identification accuracy to varying degrees, suggesting that targeting precursor emissions in these areas could enhance pollution mitigation efforts.

Indoor Scene Intelligent Identification System Based on Spatial Topological Relationship Constraints

Abstract. Indoor scene identification can provide prior environmental information for indoor positioning applications, achieving positioning enhancement. Due to the similarity of wireless signals in adjacent spaces, it can lead to incorrect identification in neighbouring scenario units. To address this problem, this paper first employs a Naive Bayes classifier to train and classify wireless signal strength data from different scenes, constructing a scene identification algorithm. Secondly, a topological relationship adjacency matrix and adjacency list tailored for indoor positioning are constructed, imposing spatial topological constraints to assist scene identification. Finally, an indoor scene intelligent identification system is developed and implemented. The experimental results indicate that the scene identification method based on spatial topological relationship constraints can effectively improve identification accuracy. The overall accuracy for 8 scenario units is 98.2%, and the identification accuracy is increased by 1.1% compared with the original method.

WiFi-Based Human Identification with Machine Learning: A Comprehensive Survey.

In the modern world of human-computer interaction, notable advancements in human identification have been achieved across fields like healthcare, academia, security, etc. Despite these advancements, challenges remain, particularly in scenarios with poor lighting, occlusion, or non-line-of-sight. To overcome these limitations, the utilization of radio frequency (RF) wireless signals, particularly wireless fidelity (WiFi), has been considered an innovative solution in recent research studies. By analyzing WiFi signal fluctuations caused by human presence, researchers have developed machine learning (ML) models that significantly improve identification accuracy. This paper conducts a comprehensive survey of recent advances and practical implementations of WiFi-based human identification. Furthermore, it covers the ML models used for human identification, system overviews, and detailed WiFi-based human identification methods. It also includes system evaluation, discussion, and future trends related to human identification. Finally, we conclude by examining the limitations of the research and discussing how researchers can shift their attention toward shaping the future trajectory of human identification through wireless signals.

Analysis and mitigation of uncertainties in damage identification by modal-curvature based methods

The objective of this paper is to address and reduce the uncertainties associated with measurement noise and discretization in damage identification methods based on modal curvature analysis. The experimental case study considers the local reduction in the bending stiffness of a slender beam under free-free and simply supported boundary conditions. First, the analysis of error sources and their propagation is theoretically set up. Second, the mitigation of uncertainties in damage localization is pursued using a two-stage approach based on multiple hypothesis testing relative to the normalized indices and the definition of a combined macro-index. Finally, the Monte Carlo method is exploited to obtain the statistical error distribution of the experimental damage position and severity predictions by randomizing the numerical displacement mode shapes with the identified noise. The present analysis allows us to find the optimal number of sensors that minimizes the combination of bias and truncation errors, to highlight how sensor spacing and data noise affect damage localization, and to determine the uncertainty bounds of the predicted damage severity. The two-stage approach, enhanced by selecting thresholds related to real noise levels and tuned on SHM objectives, appears to improve identification accuracy compared to the separate use of damage indices based on absolute confidence levels.

Modified extended Rauch–Tung–Striebel smoother method for the dynamic external excitation identification of piezoelectric vibration energy harvesting systems

A vibration energy harvester can collect vibration energy from the environment to supply low-power devices, such as wireless sensor nodes. Reducing the size and power consumption of these devices is a challenging problem. A dual-function device, which includes energy harvesting and vibration measurement, may solve this problem. Therefore, studying inversion methods for identifying the external excitation of a harvester is meaningful. In this study, a modified inversion method that combines an extended Rauch–Tung–Striebel smoother (ERTSS) with the particle swarm optimization (PSO) algorithm, i.e., the ERTSS-PSO method, is proposed to identify the time domain signal of external excitation through the voltage response of a piezoelectric vibration energy harvester. A modified ERTSS approach with sliding windows is developed. The sliding windows are partially overlapped to eliminate identification errors between the two contiguous windows. Near real-time identification is achieved for the small sliding window. PSO is utilized to estimate state noise covariance and measurement noise covariance, improving identification accuracy compared with the L-curve approach. Under this proposed framework, prior knowledge regarding the statistical data of unknown external excitations is unnecessary, enabling a more comprehensive range of applications. To assess the performance of the proposed ERTSS-PSO method, numerical simulations, including two piezoelectric vibration energy harvesting systems with third-order and fifth-order nonlinear stiffness, are conducted to verify the effectiveness of the proposed method under harmonic, multifrequency, and random excitations. The method proposed in this study is also validated by a circular laminated piezoelectric plate harvester under harmonic and random excitations in an experiment. Results from the experimental studies demonstrate that the proposed method improves identification accuracy by at least 9% compared with the extended Kalman filter, at least 7% compared with the augmented Kalman filter (AKF) with Rauch–Tung–Striebel smoother (ARTSS), and at least 40% compared with AKF under random excitation. In addition, the proposed method is unaffected by initial conditions, and it is more stable and accurate than AKF and ARTSS. The proposed method lays the theoretical foundation for identifying the external excitation of a harvester. It is also a possible solution for the miniaturization and compactness of wireless monitoring devices.

Integrative residue-intuitive machine learning and MD Approach to Unveil Allosteric Site and Mechanism for β2AR

Allosteric drugs offer a new avenue for modern drug design. However, the identification of cryptic allosteric sites presents a formidable challenge. Following the allostery nature of residue-driven conformation transition, we propose a state-of-the-art computational pipeline by developing a residue-intuitive hybrid machine learning (RHML) model coupled with molecular dynamics (MD) simulation, through which we can efficiently identify the allosteric site and allosteric modulator as well as reveal their regulation mechanism. For the clinical target β2-adrenoceptor (β2AR), we discover an additional allosteric site located around residues D792.50, F2826.44, N3187.45 and S3197.46 and one putative allosteric modulator ZINC5042. Using Molecular Mechanics/Generalized Born Surface Area (MM/GBSA) and protein structure network (PSN), the allosteric potency and regulation mechanism are probed to further improve identification accuracy. Benefiting from sufficient computational evidence, the experimental assays then validate our predicted allosteric site, negative allosteric potency and regulation pathway, showcasing the effectiveness of the identification pipeline in practice. We expect that it will be applicable to other target proteins.

Integrating Digital Imaging and AI in Forensic Odontology: Enhancing Identification Accuracy and Efficiency

Background: The integration of digital imaging and artificial intelligence (AI) in forensic odontology can significantly enhance the accuracy and efficiency of dental identification processes. These technologies provide precise and rapid analysis, which is crucial for forensic investigations. However, implementing such technologies also introduces new ethical challenges that must be addressed. Purpose: To explore the synergistic use of digital imaging and artificial intelligence (AI) in forensic odontology. This section focuses on how these technologies improve identification accuracy and efficiency while discussing the ethical considerations that arise from their implementation. Methodology: This methodology involves a comprehensive review of the existing literature on the applications of digital imaging and AI in forensic odontology. Key studies, clinical trials, and case examples are analyzed to evaluate the effectiveness of these technologies in improving forensic outcomes. The review assesses these technologies’ impact on identification accuracy and efficiency while addressing ethical concerns such as data privacy and AI biases Results: The review highlights that the combination of digital imaging and AI significantly enhances the accuracy and speed of dental identification. Advanced algorithms can process large datasets and identify subtle patterns that are often missed by human examiners, resulting in more reliable forensic outcomes. However, ethical challenges, including data privacy concerns and potential biases in AI algorithms, are critical areas that require careful consideration and regulation. Ongoing research and policy development are essential to address these issues and ensure the ethical application of AI in forensic odontology.

Improved MobileNet V3-Based Identification Method for Road Adhesion Coefficient.

To enable the timely adjustment of the control strategy of automobile active safety systems, enhance their capacity to adapt to complex working conditions, and improve driving safety, this paper introduces a new method for predicting road surface state information and recognizing road adhesion coefficients using an enhanced version of the MobileNet V3 model. On one hand, the Squeeze-and-Excitation (SE) is replaced by the Convolutional Block Attention Module (CBAM). It can enhance the extraction of features effectively by considering both spatial and channel dimensions. On the other hand, the cross-entropy loss function is replaced by the Bias Loss function. It can reduce the random prediction problem occurring in the optimization process to improve identification accuracy. Finally, the proposed method is evaluated in an experiment with a four-wheel-drive ROS robot platform. Results indicate that a classification precision of 95.53% is achieved, which is higher than existing road adhesion coefficient identification methods.

Sea Fog Recognition near Coastline Using Millimeter-Wave Radar Based on Machine Learning

Sea fog is a hazardous natural phenomenon that reduces visibility, posing a threat to ports and nearshore navigation, making the identification of nearshore sea fog crucial. Millimeter-wave radar has significant advantages over satellites in capturing sudden and localized sea fog weather. The use of millimeter-wave radar for sea fog identification is still in the exploratory stage in operational fields. Therefore, this paper proposes a nearshore sea fog identification algorithm that combines millimeter-wave radar with multiple machine learning methods. Firstly, Density-Based Spatial Clustering of Applications with Noise (DBSCAN) is used to partition radar echoes, followed by the K-means clustering algorithm (KMEANS) to divide the partitions into recognition units. Then, Sea-Fog-Recognition-Convolutional Neural Network (SFRCNN) is used to classify whether the recognition units are sea fog areas, and finally, the partition coverage algorithm is employed to improve identification accuracy. The experiments conducted using millimeter-wave radar observation data from the Pingtan Meteorological Observation Base in Fujian, China, achieved an identification accuracy of 96.94%. The results indicate that the proposed algorithm performs well and expands the application prospects of such equipment in meteorological operations.

Bayesian-guided operational modal identification of a highway bridge considering non-uniform sampling

During the structural health monitoring of bridges, it has been observed that the vibration data collected can sometimes be randomly lost or sampled non-uniformly. This leads to a low signal-to-noise ratio in the spectral functions of the measured data, making it difficult to identify weak modes. To address this issue, a framework for operational modal identification is proposed in this study. It utilizes the fast Bayesian fast Fourier transform (FFT) method to estimate the modal parameters of highway bridges considering the non-uniform monitoring data. The initial frequency parameters for the fast Bayesian FFT approach are automatically determined using the proposed autoregressive (AR) power spectral density (PSD)-guided peak picking method. This overcomes the challenge of capturing initial frequencies related to weakly contributed modes. Additionally, the bandwidth parameter for each mode is determined using the modal assurance criterion (MAC) of the first left singular vectors of PSD matrices. Furthermore, when analyzing non-uniform vibration data, it is recommended to use the non-uniform FFT (NUFFT) for calculating PSD functions in order to improve identification accuracy. The proposed method is validated using acceleration data from both a numerical model and a real-world bridge. The results demonstrate that the identification uncertainty of modal parameters increases with higher non-uniform levels.

Honeysuckle flower stage identification based on improved YOLOv5s

AbstractThe medicinal constituents of Chinese herbal medicine honeysuckle (Lonicera japonica Thunb) vary at different flower stages. In order to ensure that the medicinal value is maximized, it is necessary to identify its flower stage before harvesting. However, at present, this study can only be accomplished by manual visual recognition, which is inefficient and costly. Therefore, there is an urgent need to develop an automatic detection technique with high maturity, fast detection speed, and strong model deployment capability. In order to adapt to the problems of different flower size and color texture similarity and complex background, this study chooses YOLOv5s algorithm for adaptive modification. First, a small detection layer is added to the network to enhance feature extraction and improve the accuracy of identifying small honeysuckle. Second, attention mechanism is incorporated into the backbone network to suppress background interference and improve identification accuracy. Finally, the original IoU‐NMS is replaced by the DIoU‐NMS algorithm, which improves the bounding box regression rate while reducing the leakage rate when overlapping or occluded. The test results showed that the P was increased from 80.0% to 92.7%, the R was increased from 78.6% to 80.2%, and the mean average precision was increased from 86.2% to 90.6%. Furthermore, the model was verified at both long range and short range, and the tests data indicate that the identification accuracy was no less than 90% in 3 m without serious occlusion. This study laid a solid foundation for accurate honeysuckle flower stage identification and provided technical support for real‐time machine picking honeysuckle.

Identification study of soil types based on feature factors of XRF spectrum combining with machine learning

Soil type significantly influences the detection accuracy of heavy metals using X-ray fluorescence (XRF) technology. Rapid and accurate soil type identification is crucial for selecting appropriate XRF quantitative analysis methods for soil heavy metals, thereby enhancing analysis accuracy. This study utilized 26 soil samples from 10 distinct soil types, extracting 13 feature factors for soil type identification by analyzing XRF spectral variability. These factors were then integrated with three machine learning methods: Random Forest (RF), Support Vector Machine (SVM), and Backpropagation Neural Network (BPNN). The effectiveness of these methods in soil type identification was compared, highlighting the importance of XRF spectral feature factor extraction. The results demonstrate that identification based on feature factor extraction from XRF spectral variability markedly improves identification accuracy, stability and speed compared to full-spectrum XRF analysis. When identifying soil types by the gross area of spectral peaks of XRF feature factors, the accuracies of three machine learning methods—RF, SVM, and BPNN—were 99.62%, 99.04%, and 98.85%, respectively. Random Forest achieved the highest accuracy (99.62%) and fastest operation speed (0.179 s). Therefore, by extracting the differential features of XRF spectra and combining them with machine learning methods, it is possible to quickly and accurately recognize and judge soil types. This study demonstrates the successful and accurate identification of soil types using machine learning combined with XRF spectroscopy.It establishes an important methodological foundation for the future development of fast and accurate field testing equipment for soil heavy metals using XRF technology.

Identification of Missing Person using CNN

Our project aims to leverage Convolutional Neural Networks (CNNs) for the identification of missing persons. CNNs, a class of deep learning algorithms widely used in image recognition tasks, offer promising potential in automating and enhancing the identification process. The project aims to develop a robust system using CNN models to match unidentified individuals with missing person databases, improving identification accuracy and providing closure. The proposed approach demonstrates potential in assisting law enforcement agencies and missing persons organizations by providing a reliable and efficient means of cross-referencing images from various sources, such as surveillance footage, social media, and public records. Additionally, the flexibility of CNNs allows for the integration of other biometric markers, including fingerprints and voice recognition, to increase the accuracy and reliability of identifications. This research underscores the importance of using artificial intelligence and machine learning in social good applications, highlighting the potential for technology to play a transformative role in reuniting families and bringing closure to unresolved cases. Future work will focus on refining the model, enhancing privacy protections, and ensuring ethical use in real-world applications.

Boosting person ReID feature extraction via dynamic convolution

Extraction of discriminative features is crucial in person re-identification (ReID) which aims to match a query image of a person to her/his images, captured by different cameras. The conventional deep feature extraction methods on ReID employ CNNs with static convolutional kernels, where the kernel parameters are optimized during the training and remain constant in the inference. This approach limits the network's ability to model complex contents and decreases performance, particularly when dealing with occlusions or pose changes. In this work, to improve the performance without a significant increase in parameter size, we present a novel approach by utilizing a channel fusion-based dynamic convolution backbone network, which enables the kernels to change adaptively based on the input image, within two existing ReID network architectures. We replace the backbone network of two ReID methods to investigate the effect of dynamic convolution on both simple and complex networks. The first one called Baseline, is a simpler network with fewer layers, while the second, CaceNet represents a more complex architecture with higher performance. Evaluation results demonstrate that both of the designed dynamic networks improve identification accuracy compared to the static counterparts. A significant increase in accuracy is reported under occlusion tested on Occluded-DukeMTMC. Moreover, our approach achieves a performance comparable to the state-of-the-art on Market1501, DukeMTMC-reID, and CUHK03 with a limited computational load. These findings validate the effectiveness of the dynamic convolution in enhancing the person ReID networks and push the boundaries of performance in this domain.

Cross-Social-Network User Identification Based on Bidirectional GCN and MNF-UI Models

Due to the distinct functionalities of various social network platforms, users often register accounts on different platforms, posing significant challenges for unified user management. However, current multi-social-network user identification algorithms heavily rely on user attributes and cannot perform user identification across multiple social networks. To address these issues, this paper proposes two identity recognition models. The first model is a cross-social-network user identification model based on bidirectional GCN. It calculates user intimacy using the Jaccard similarity coefficient and constructs an adjacency matrix to accurately represent user relationships in the social network. It then extracts cross-social-network user information to accomplish user identification tasks. The second model is the multi-network feature user identification (MNF-UI) model, which introduces the concept of network feature vectors. It effectively maps the structural features of different social networks and performs user identification based on the common features of seed nodes in the cross-network environment. Experimental results demonstrate that the bidirectional GCN model significantly outperforms baseline algorithms in cross-social-network user identification tasks. The MNF-UI (multi-network feature user identification) model can operate in situations with two or more networks with inconsistent structures, resulting in improved identification accuracy. These two user identification algorithms provide technical and theoretical support for in-depth research on social network information integration and network security maintenance.

Application of multi-temporal satellite imagery for urban tree species identification

Accurate tree inventories are critical for urban forest management but challenging to obtain, as many urban trees are on private property (backyards, etc.) and are excluded from public inventories. Here, we examined the feasibility of tree species identification in a large heterogenous urban area (>850 km2) by using multi-temporal PlanetScope images (3.2 m resolution, multi-spectral) and inventory data from more than 20,000 ground observations within the urban forest of the Greater Chicago area. Our approach achieved an overall classification accuracy of 0.60 and 0.71 for 18 species and ten genera, respectively, but varied from moderate to high for certain species (0.59–0.92) and genera (0.61–0.91). In particular, we identified key host tree species (Fraxinus americana, F. pennsylvanica, and Acer saccharinum) for two damaging invasive insects, emerald ash borer (EAB, Agrilus planipennis) and Asian longhorn beetle (ALB, Anoplophora glabripennis), with over 0.80 accuracies. In addition, we demonstrated that including images from the autumn months (September–November), either for a single-season model or a combined multiple-season model, improved the identification accuracy of temperate deciduous trees. Further, the high classification accuracy of support vector machine (SVM) over random forest (RF) and neural network (NN) approaches suggests that future work might benefit from comparing multiple classification methods to select the approach that maximizes species classification accuracy. Our study demonstrated the potential for applying multi-temporal high-resolution images in urban tree classification, which can be used for urban forest management at a large spatial scale.

Strategy to expedite gas sensor calibration based on cyclic temperature operation and data augmentation

The gas sensor calibration using data-driven approach requires a large volume of training data. The conventional method demands numerous measurements of response data, consuming substantial manpower and time resources. To expedite and streamline the sensor calibration and model deployment, this study proposes a cost-effective and practical calibration strategy combining data augmentation and data preprocessing. This strategy aims to minimize both the collection time and amount of measurement data necessary while ensuring high identification accuracy and generalization performance. The random cropping technique based on sliding sampling window is developed to expand the sample size by randomly and consecutively resampling new data from the limited raw response data. This approach allows any fixed-length sequence from the raw response curve to serve as the training and test sample, facilitating fast and real-time measurement for online detection and dynamic analysis. Furthermore, data preprocessing techniques using sequence normalization and fast Fourier transform are employed to extract the species and concentration features, thereby mitigating drift-like effects in the response data. The effectiveness of this strategy is demonstrated with real measurement data, showcasing significant improvements in identification accuracy and generalization performance compared to conventional methods.

The Effect of Visual Articulatory Cues on the Identification of Mandarin Tones by Children With Cochlear Implants.

This study explored the facilitatory effect of visual articulatory cues on the identification of Mandarin lexical tones by children with cochlear implants (CIs) in both quiet and noisy environments. It also explored whether early implantation is associated with better use of visual cues in tonal identification. Participants included 106 children with CIs and 100 normal-hearing (NH) controls. A tonal identification task was employed using a two-alternative forced-choice picture-pointing paradigm. Participants' tonal identification accuracies were compared between audio-only (AO) and audiovisual (AV) modalities. Correlations between implantation ages and visual benefits (accuracy differences between AO and AV modalities) were also examined. Children with CIs demonstrated an improved identification accuracy from AO to AV modalities in the noisy environment. Additionally, earlier implantation was significantly correlated with a greater visual benefit in noise. These findings indicated that children with CIs benefited from visual cues on tonal identification in noise, and early implantation enhanced the visual benefit. These results thus have practical implications on tonal perception interventions for Mandarin-speaking children with CIs.

Geometric error modeling and decoupling identification of rotary axis of five-axis machine tool based on spatial trajectory planning

This paper proposes a method for identification of geometric error and the improvement of measuring accuracy of rotary axes. The primary objective of this work concentrates on the decoupling identify of geometric errors of rotary axis, and the optimal selection of measurement parameters (velocity and measuring points), which can ensure measurement completeness and improve accuracy simultaneously. The proposed approach involves the construction of spatial trajectory driven four-axis motion synchronously. Subsequently, measurement parameters are optimal selected with identification matrix. In the next step, PDGEs are represented using NURBS rapidly solved. The investigations reveal that error decoupling can further enhance compensation accuracy. It was also shown that simply increasing the number of measurements does not improve identification accuracy. The result of experiment indicates that the average compensation rate of trajectory reaches 59.73 %. Therefore, the proposed method can be employed in the reliable calibration of geometric errors particular to five-axis machine tool.

Characterization and recognition of three-dimensional excitation-emission matrix spectra of wastewater from six typical categories

In this study, we analyzed the characteristics of three-dimensional excitation-emission matrix spectra (EEMs) of 150 samples from five industrial wastewater types and domestic sewage to track water pollution sources effectively. We then developed a recognition model for wastewater EEMs by establishing a feature dataset containing fluorescence peak values and parameters derived from EEMs, integrated with machine learning techniques. This model enables the rapid and precise identification of pollution sources. Our findings suggest that although the EEMs of the six wastewater categories are distinct, visual differentiation is challenging. This was confirmed by cosine similarity assessments, showing some samples with low within-group (< 0.8) and high between-group (> 0.95) similarities. Despite significant variations in EEMs features across wastewater categories, identifying specific pollutants remains difficult, especially for pulp mills and leather effluents. Among the tested classification algorithms, Support Vector Machine (SVM) achieved the highest performance with 91.7 % accuracy, 94 % precision, 91 % recall, and 92 % F1-score, outperforming K-Nearest Neighbors and Partial Least Squares Discriminant Analysis. The SVM significantly improved identification accuracy for pulp mill and leather processing wastewaters compared to other models. To enhance identification accuracy, further exploration of EEMs features and expanding the training dataset are recommended. Combining EEMs features with machine learning presents a promising method for improving water pollution supervision and source tracing in environmental management practices.