Matching Accuracy Research Articles

Tightly Coupled SLAM Algorithm Based on Similarity Detection Using LiDAR-IMU Sensor Fusion for Autonomous Navigation

In recent years, the rise of unmanned technology has made Simultaneous Localization and Mapping (SLAM) algorithms a focal point of research in the field of robotics. SLAM algorithms are primarily categorized into visual SLAM and laser SLAM, based on the type of external sensors employed. Laser SLAM algorithms have become essential in robotics and autonomous driving due to their insensitivity to lighting conditions, precise distance measurements, and ease of generating navigation maps. Throughout the development of SLAM technology, numerous effective algorithms have been introduced. However, existing algorithms still encounter challenges, such as localization errors and suboptimal utilization of sensor data. To address these issues, this paper proposes a tightly coupled SLAM algorithm based on similarity detection. The algorithm integrates Inertial Measurement Unit (IMU) and LiDAR odometry modules, employs a tightly coupled processing approach for sensor data, and utilizes curvature feature optimization extraction methods to enhance the accuracy and robustness of inter-frame matching. Additionally, the algorithm incorporates a local keyframe sliding window method and introduces a similarity detection mechanism, which reduces the real-time computational load and improves efficiency. Experimental results demonstrate that the algorithm achieves superior performance, with reduced positioning errors and enhanced global consistency, in tests conducted on the KITTI dataset. The accuracy of the real trajectory data compared to the ground truth is evaluated using metrics such as ATE (absolute trajectory error) and RMSE (root mean square error).

Improving the position accuracy via merged line and point features for GNSS/SINS/Vision integration in a degradation environment

Abstract In complex urban environments, the GNSS/INS integration diverges over time when severe GNSS degradation occurs due to the harsh environment. Based on visual point features, the visual-inertial odometry (VIO) system has been integrated with the GNSS to solve this problem. However, the accuracy of this system decreases when the GNSS and visual point features degrade simultaneously in the GNSS harsh environments and low-texture scenes. The artificial structures in complex urban environments render line features more reliable than point features. To further improve the positioning accuracy of the point-based fusion system, we propose a GNSS/INS/Vision integration method based on point and line features, where visual measurements in a sliding window and the GNSS position are used to reduce or eliminate the cumulative errors of the IMU state in the back end. In addition, visual features are merged at the front end via an adaptive threshold based on the line feature length. Two urban infield experiments are used to evaluate the effectiveness of the proposed method through improvements in the time consumption of line feature matching and position accuracy. The analysis demonstrates that the average time consumption is reduced by 50.03% and 82.19%, and the maximum position error is reduced by 42.56% and 55.48%, respectively in the two scenarios. Additional evaluations based on KAIST Urban 38 public datasets indicate that our proposed system achieves a better positioning accuracy than that of VINS-Fusion, which is a state-of-the-art GNSS/INS/Vision integration system that uses only point features.

Design of a Student Recommendation Platform Based on Learning Behavior and Habit Training

This study innovatively built an intelligent analysis platform for learning behavior, which deeply integrated the cutting-edge technology of big data and Artificial Intelligence (AI), \mined and analyzed students’ learning data, and realized the personalized customization of learning resources and the accurate matching of intelligent learning partners. With the help of advanced algorithms and multi-dimensional data fusion strategies, the platform not only promotes positive interaction and collaboration in the learning environment but also provides teachers with comprehensive and in-depth students’ learning portraits, which provides solid support for the implementation of precision education and the personalized adjustment of teaching strategies. In this study, a recommender system based on user similarity evaluation and a collaborative filtering mechanism is carefully designed, and its technical architecture and implementation process are described in detail.

Global Context Volume construction and semantics-guided disparity refinement for stereo matching

The accuracy of stereo matching has been greatly improved with the advent of convolutional neural networks. However, existing methods still perform poorly in regions of low-texture and reflections due to insufficient matching information. In this paper, we propose a novel semantic stereo matching network named GSSNet, which combines global context information with high-level semantic clues to further enrich the matching features. Specifically, we employ Swin Transformer as a joint feature extractor to capture multi-scale global features for both semantic segmentation and disparity estimation. By fusing low-resolution multi-scale feature maps progressively, we construct a Global Context Volume (GCV) representation covering long-range receptive field for initial disparity estimation. Initial disparities are further refined with a high-level embedding learned from the semantic segmentation branch of the network through residual learning. During training, due to the lack of joint annotations, we also propose a simple and effective pseudo-labeling strategy. Comprehensive experimental results demonstrated on various datasets manifest the effectiveness of the proposed GSSNet. Among all published methods as of 9, February 2024, our approach ranks 1st on KITTI 2012 leaderboard and 3rd on KITTI 2015 leaderboard, and produces competitive results on Scene Flow and Cityscapes datasets. Code will be available at https://github.com/Twil-7/GSSNet.

Deep ensemble architectures with heterogeneous approach for an efficient content-based image retrieval

<span lang="EN-US">In the field of digital image processing, content-based image retrieval (CBIR) has become essential for searching images based on visual content characteristics like color, shape, and texture, rather than relying on text-based annotations. To address the increasing demands for efficiency and precision in CBIR systems, we introduce the HybridEnsembleNet methodology. HybridEnsembleNet combines deep learning algorithms with an asymmetric retrieval framework to optimize feature extraction and comparison in extensive image databases. This novel approach, specifically custom-made for CBIR, employs a lightweight query structure skilled at handling large-scale data under resource-constrained environments. The experiments were performed on the ROxford and RParis datasets. The deep learning component of HybridEnsembleNet significantly refines the accuracy of image matching and retrieval. RParis The ROxford dataset, specifically in the medium and hard difficulty benchmarks, demonstrates an enhancement of 5.53% and 10.44%, respectively. Similarly, the RParis dataset, under medium and hard benchmarks, exhibits improvements of 3.01% and 5.83%, showcasing superior performance compared to existing models. By overcoming the traditional limitations of CBIR systems in mean average precision (mAP) metrics, HybridEnsembleNet provides a scalable, efficient, and more accurate solution for retrieving relevant images from vast digital libraries.</span>

Selective weighted least square and piecewise bilinear transformation for accurate satellite DSM generation

One of the main products of multi-view stereo (MVS) high-resolution satellite (HRS) images in photogrammetry and remote sensing is digital surface model (DSM). Producing DSMs from MVS HRS images still faces serious challenges due to various reasons such as complexity of imaging geometry and exterior orientation model in HRS, as well as large dimensions and various geometric and illumination variations. The main motivation for conducting this research is to provide a novel and efficient method that enhances the accuracy and completeness of extracting DSM from HRS images compared to existing recent methods. The proposed method called Sat-DSM, consists of five main stages. Initially, a very dense set of tie-points is extracted from the images using a tile-based matching method, phase congruency-based feature detectors and descriptors, and a local geometric consistency correspondence method. Then, the process of Rational Polynomial Coefficients (RPC) block adjustment is performed to compensate the RPC bias errors. After that, a dense matching process is performed to generate 3D point clouds for each pair of input HRS images using a new geometric transformation called PWB (pricewise bilinear) and an accurate area-based matching method called SWLSM (selective weighted least square matching). The key innovations of this research include the introduction of SWLSM and PWB methods for an accurate dense matching process. The PWB is a novel and simple piecewise geometric transformation model based on superpixel over-segmentation that has been proposed for accurate registration of each pair of HRS images. The SWLSM matching method is based on phase congruency measure and a selection strategy to improve the well-known LSM (least square matching) performance. After dense matching process, the final stage is spatial intersection to generate 3D point clouds, followed by elevation interpolation to produce DSM. To evaluate the Sat-DSM method, 12 sets of MVS-HRS data from IRS-P5, ZY3-1, ZY3-2, and Worldview-3 sensors were selected from areas with different landscapes such as urban, mountainous, and agricultural areas. The results indicate the superiority of the proposed Sat-DSM method over four other methods CATALYST, SGM (Semi-global matching), SS-DSM (structural similarity based DSM extraction), and Sat-MVSF in terms of completeness, RMSE, and MEE. The demo code is available at https://www.researchgate.net/publication/377721674_SatDSM.

Detective Gadget: Generic Iterative Entity Resolution over Dirty Data

In the era of Big Data, entity resolution (ER), i.e., the process of identifying which records refer to the same entity in the real world, plays a critical role in data-integration tasks, especially in mission-critical applications where accuracy is mandatory, since we want to avoid integrating different entities or missing matches. However, existing approaches struggle with the challenges posed by rapidly changing data and the presence of dirtiness, which requires an iterative refinement during the time. We present Detective Gadget, a novel system for iterative ER that seamlessly integrates data-cleaning into the ER workflow. Detective Gadgetemploys an alias-based hashing mechanism for fast and scalable matching, check functions to detect and correct mismatches, and a human-in-the-loop framework to refine results through expert feedback. The system iteratively improves data quality and matching accuracy by leveraging evidence from both automated and manual decisions. Extensive experiments across diverse real-world scenarios demonstrate its effectiveness, achieving high accuracy and efficiency while adapting to evolving datasets.

IDPriU: A two-party ID-private data union protocol for privacy-preserving machine learning

Due to significant data security concerns in machine learning, such as the data silo problem, there has been a growing trend towards the development of privacy-preserving machine learning applications. The initial step in training data across silos involves establishing secure data joins, specifically private data joins, to ensure the consistency and accuracy of the dataset. While the majority of current research focuses on the inner join of private data, this paper specifically addresses the privacy-preserving full join of private data and develops two-party unbalanced private data full join protocols utilizing secure multi-party computation tools. Notably, our paper introduces the novel component of Private Match-and-Connect (PMC), which performs a union operation on the ID and feature values, and ensure the secret sharing of the resulting union set. Each participant receives only a portion of the secret share, thereby guaranteeing data security during the pre-processing phase. Furthermore, we propose the two-party ID-private data union protocol (IDPriU), which facilitates secure and accurate matching of feature value shares and ID shares and also enables the data alignment. Our protocol represents a significant advancement in the field of privacy-preserving data preprocessing in machine learning and privacy-preserving federated queries. It extends the concept that private data joins are limited to inner connections, offering a novel approach by Private Set Union (PSU). We have experimentally implemented our protocol and obtained favorable results in terms of both runtime and communication overhead.

New Insight of Nanosheet Enhanced Oil Recovery Modeling: Structural Disjoining Pressure and Profile Control Technique Simulation

This study presents a novel Enhanced Oil Recovery (EOR) method using Smart Black Nanocards (SLNs) to mitigate the environmental impact of conventional thermal recovery, especially under global warming. Unlike prior studies focusing on wettability alteration via adsorption, this research innovatively models ‘oil film detachment’ in a reservoir simulator to achieve wettability alteration. Using the CMG-STARS (2020) simulator, this study highlights SLNs’ superior performance over traditional chemical EOR and spherical nanoparticles by reducing residual oil saturation and shifting wettability toward water-wet conditions. The structural disjoining pressure (SDP) of SLNs reaches 20.99 × 103 Pa, 16.5 times higher than spherical particles with an 18.5 nm diameter. Supported by the Percus–Yevick (PY) theory, the numerical model achieves high accuracy in production history matching, with oil recovery and water cut fitting within precision error ranges of 0.02 and 0.05, respectively. This research advances chemical EOR technologies and offers an environmentally sustainable, efficient recovery strategy for low-permeability and heavy oil reservoirs, serving as a promising alternative to thermal methods.

Relation Constrained Capsule Graph Neural Networks for Non-Rigid Shape Correspondence

Non-rigid 3D shape correspondence aims to establish dense correspondences between two non-rigidly deformed 3D shapes. However, the variability and symmetry of non-rigid shapes usually lead to mismatches due to shape deformation, topological changes, or data with severe noise. To finding an accurate correspondence between 3D dynamic shapes for the local deformation complexity, this article proposes a Relation Constrained Capsule Graph Network (RC-CGNet), which combines global and local features by encouraging the relation constraints between the embedding feature space and the input shape space based on the functional maps framework. Specifically, we design a Diffusion Graph Attention Network (DGANet) to segment the surface into parts with correct edge boundary between two regions. The Minimum Spanning Tree (MST) of geodesic curves among the singularities obtained from the segmented parts is added as relation constraints, which can compute isometric correspondences in both direct and symmetric directions. Besides that, the relation-and-attention constrained neural networks are designed to learn the shape correspondence via attention-aware CapsNet and functional maps under relation constraints. To improve the convergence speed and matching accuracy, we propose an optimized residual network structure based on the Nesterov Accelerated Gradient (NAG) to extract local features, and use graph convolution structure to extract global features. Moreover, a lightweight Gated Attention Module (GAM) is designed to fuse global and local features to obtain a richer feature representation. Since the capsule network has better spatial reasoning ability than the traditional convolutional neural network, our novel network architecture is a dual-route capsule network based on Routing Attention Fusion Block (RAFB), filtering low-discriminative capsules from a holistic view by exploiting geometric hierarchical relationships of semantic parts. Experiments on open datasets show that our method has excellent accuracy and wide adaptability.

ZTF SN Ia DR2: The spectral diversity of Type Ia supernovae in a volume-limited sample

More than 3000 spectroscopically confirmed Type Ia supernovae (SNe Ia) are presented in the second data release (DR2) of the Zwicky Transient Facility survey. In this paper we detail the spectral properties of 482 SNe Ia near maximum light, up to a redshift limit of z leq 0.06. We measured the velocities and pseudo-equivalent widths (pEW) of key spectral features ( and and investigated the relation between the properties of the spectral features and the photometric properties from the SALT2 light-curve parameters as a function of spectroscopic sub-class. We discuss the non-negligible impact of host galaxy contamination on SN Ia spectral classifications, and we investigate the accuracy of spectral template matching of the DR2 sample. We define a new subclass of underluminous SNe Ia (04gs-like) that lie spectroscopically between normal SNe Ia and transitional 86G-like SNe Ia (stronger than normal SNe Ia, but significantly weaker features than 86G-like SNe). We model these 04gs-like SN Ia spectra using the radiative-transfer spectral synthesis code tardis and show that cooler temperatures alone are unable to explain their spectra; some changes in elemental abundances are also required. However, the broad continuity in spectral properties seen from bright (91T-like) to faint normal SN Ia, including the transitional and 91bg-like SNe Ia, suggests that variations within a single explosion model may be able to explain their behaviour.

Precise wavelength alignment of second-harmonic generation in thin-film lithium niobate resonators.

Second-harmonic generation (SHG) plays a significant role in modern photonic technology. Integrated photonic resonators fabricated with thin-film lithium niobate can achieve ultrahigh efficiencies by combining small mode volumes with high material nonlinearity. Cavity-enhanced SHG requires accurate phase and frequency matching conditions, where fundamental and second-harmonic wavelengths are both on resonance. However, this double-resonance condition can typically be realized only at a fixed random wavelength due to the high sensitivity of photonic resonances to the device geometry and fabrication variations. Here, we propose a novel method that can achieve the double-resonance condition over a large wavelength range. We combine thermal-optic and electro-optic (EO) effects to realize the separate tuning of fundamental and second-harmonic resonances. We demonstrated that the optimum SHG efficiency can be maintained over a wavelength range that exceeds the limit achievable with only thermal tuning. With this flexible tuning capability, we further show the precise alignment of SHG wavelengths of two separate thin-film lithium niobate resonators without sacrificing efficiencies.

Medial Skeletal Diagram: A Generalized Medial Axis Approach for Compact 3D Shape Representation

We propose the Medial Skeletal Diagram, a novel skeletal representation that tackles the prevailing issues around skeleton sparsity and reconstruction accuracy in existing skeletal representations. Our approach augments the continuous elements in the medial axis representation to effectively shift the complexity away from the discrete elements. To that end, we introduce generalized enveloping primitives, an enhancement over the standard primitives in the medial axis, which ensure efficient coverage of intricate local features of the input shape and substantially reduce the number of discrete elements required. Moreover, we present a computational framework for constructing a medial skeletal diagram from an arbitrary closed manifold mesh. Our optimization pipeline ensures that the resulting medial skeletal diagram comprehensively covers the input shape with the fewest primitives. Additionally, each optimized primitive undergoes a post-refinement process to guarantee an accurate match with the source mesh in both geometry and tessellation. We validate our approach on a comprehensive benchmark of 100 shapes, demonstrating the sparsity of the discrete elements and superior reconstruction accuracy across a variety of cases. Finally, we exemplify the versatility of our representation in downstream applications such as shape generation, mesh decomposition, shape optimization, mesh alignment, mesh compression, and user-interactive design.

The Ground-Penetrating Radar Image Matching Method Based on Central Dense Structure Context Features

Subsurface structural distribution can be detected using Ground-Penetrating Radar (GPR). The distribution can be considered as road fingerprints for vehicle positioning. Similar to the principle of visual image matching for localization, the position coordinates of the vehicle can be calculated by matching real-time GPR images with pre-constructed reference GPR images. However, GPR images, due to their low resolution, cannot extract well-defined geometric features such as corners and lines. Thus, traditional visual image processing algorithms perform inadequately when applied to GPR image matching. To address this issue, this paper innovatively proposes a GPR image matching and localization method based on a novel feature descriptor, termed as central dense structure context (CDSC) features. The algorithm utilizes the strip-like elements in GPR images to improve the accuracy of GPR image matching. First, a CDSC feature descriptor is designed. By applying threshold segmentation and extremum point extraction to the GPR image, stratified strip-like elements and pseudo-corner points are obtained. The pseudo-corner points are treated as the centers, and the surrounding strip-like elements are described in context to form the GPR feature descriptors. Then, based on the feature description method, feature descriptors for both the real-time image and the reference image are calculated separately. By searching for the nearest matching point pairs and removing erroneous pairs, GPR image matching and localization are achieved. The proposed algorithm was evaluated on datasets collected from urban roads and railway tracks, achieving localization errors of 0.06 m (RMSE) and 1.22 m (RMSE), respectively. Compared to the traditional Speeded Up Robust Features (SURF) visual image matching algorithm, localization errors were reduced by 86.6% and 95.7% in urban road and railway track scenarios, respectively.

Cross-Modal Contrastive Learning with a Style-Mixed Bridge for Single Image 3D Shape Retrieval

Image-based 3D shape retrieval (IBSR) is a cross-modal matching task which searches similar shapes from a 3D repository using a natural image. Continuous attention has been paid to this topic, such as joint embedding, adversarial learning, and contrastive learning. Modality gap and diversity of instance similarities are two obstacles for accurate and fine-grained cross-modal matching. To overcome the two obstacles, we propose a style-mixed contrastive learning method (SC-IBSR). On one hand, we propose a style transition module to mix the styles of images and rendered shape views to form an intermediate style and inject it into image contents. The obtained style-mixed image features serve as a bridge for later contrastive learning in order to alleviate the modality gap. On the other hand, the proposed strategy of fine-grained consistency constraint aims at cross-domain contrast and considers the different importance of negative (positive) samples. Extensive experiments demonstrate the superiority of the style-mixed cross-modal contrastive learning on both the instance-level retrieval benchmark (i.e., Pix3D, Stanford Cars, and Comp Cars that annotate shapes to images), and the unsupervised category-level retrieval benchmark (i.e., MI3DOR-1 and MI3DOR-2 with unlabeled 3D shapes). Moreover, experiments are conducted on Office-31 dataset to validate the generalization capability of our method. Code and pretrained models will be available at https://github.com/honoria0204/SC-IBSR .

A Similar Feature Point Matching Method for Aerial Electric Power Tower Images Based on a One-Ring Neighborhood of Vertices

Due to the susceptibility of images to various factors such as scale changes, imaging conditions, and image noise, traditional feature point matching methods are difficult to achieve ideal matching accuracy, leading to many challenges in the automatic analysis and processing of power tower images. To improve the matching accuracy of similar feature points in aerial images of electric power tower, this study proposes a method for matching similar feature points in aerial images of electric power tower based on a vertex ring neighborhood, addressing the problem of large matching errors caused by factors such as scale and imaging conditions. This method first adopts the feature point extraction technique of electric power tower image based on decomposition and filtering, and constructs the Laplacian pyramid of the original image. Subsequently, filter banks are used to decompose the pyramid image in different directions, extract local extremum points as candidate feature point sets, and merge them to obtain the final feature point set. On this basis, taking into account the spatial relationships and geometric characteristics around the feature points, a texture mapping algorithm based on vertex ring neighborhood and patch classification is applied to construct a vertex ring neighborhood structure and extract geometrically representative electric power tower features. Finally, by using a texture feature point matching method based on the principle of similarity, the similarity between the real-time image and the reference image features is calculated for matching, and combined with the Babbitt coefficient to remove mismatched feature points, accurate matching of similar feature points in aerial electric power tower images is achieved. The experimental results show that this method has a mean square error of less than 0.1Pixel in matching similar texture feature points of aerial power tower images under various working conditions, significantly improving the matching accuracy and providing an effective tool for automatic analysis and processing of power tower images.

Bearing life prediction method based on novel health indicators and improved similarity curve matching

Abstract To tackle the issues of low accuracy in similarity curve matching and the challenges in measuring the similarity of degraded curve segments in rolling bearing life prediction methods, we propose a novel bearing remaining useful life (RUL) prediction method utilizing new health indicators and improved similarity curve matching techniques. Initially, time and frequency domain features are extracted from the bearing vibration signals. Subsequently, the multi-dimensional features are meticulously screened using monotonicity, Spearman's correlation, and uncertainty metrics to construct the health indicator (HI) and the health indicator library for the bearing. Subsequently, the support vector data description (SVDD) combined with the PID search algorithm is employed to ascertain the initial prediction time of the bearing and to construct the corresponding health indicator library. Following this, both global and local feature capture are conducted using dual dynamic time warping (DDTW) in conjunction with a sliding window to predict the remaining service life of rolling bearings.
The proposed method is validated on the PHM2012 and XJTU-SY datasets and benchmarked against the latest research. The results demonstrate that our method significantly enhances the prediction accuracy of the remaining useful life of bearings, underscoring the effectiveness and superiority of the proposed approach.and superiority of the proposed approach.

Image–Text Matching Model Based on CLIP Bimodal Encoding

Image–text matching is a fundamental task in the multimodal research field, connecting computer vision and natural language processing by aligning visual content with corresponding textual descriptions. Accurate matching is critical for applications such as image captioning and text-based image retrieval yet remains challenging due to the differences in data modalities. This paper addresses these challenges by proposing a robust image–text matching model inspired by Contrastive Language–Image Pre-training (CLIP). Our approach employs the Vision Transformer (ViT) model as the image encoder and Bidirectional Encoder Representations from Transformers (Bert) as the text encoder, integrating these into a shared vector space to measure semantic similarity. We enhance the model’s training efficiency using the LiT-tuning paradigm to optimize learning through a cosine decay strategy for dynamic adjustment of the learning rate. We validate our method on two benchmark datasets, WuKong and Flickr30k, demonstrating that our model achieves superior performance and significantly improves key evaluation metrics. The results underscore the model’s effectiveness in achieving accurate and robust image–text alignment.

Robust Multimodal Image Matching Based on Radiation Invariant Phase Correlation

Abstract. Due to the influence of nonlinear radiation distortion and geometric deformation, achieving multimodal image matching remains a challenging task. To address these issues, this paper proposes a method called radiation invariant phase correlation (RIPC) to simultaneously estimate the rotation, scale, and displacement changes of multimodal image pairs. Firstly, based on the local structure characteristics of the image itself, we harness the nonlinear invariance of kernel canonical correlation analysis to devise the multimodal local self-correlation (MLSC) descriptor. This descriptor is resilient to nonlinear radiative differences, as well as local rotation and scale variations. Subsequently, we incorporate the log-polar coordinate transformation to capture the overall rotation and scale changes in the image, enabling independent representation of these factors on the Cartesian coordinate system. Finally, drawing upon the continuity of displacement estimation, as well as rotation and scale estimation, we construct a five-dimensional descriptor tailored for phase correlation. Extensive experiments conducted on five open-source datasets demonstrate that our proposed method surpasses state-of-the-art (SOTA) techniques in matching performance. Furthermore, our RIPC method achieves matching accuracy within 2-pixel threshold, which underscores its effectiveness in multimodal remote sensing image matching.

Enhancing human phenotype ontology term extraction through synthetic case reports and embedding-based retrieval: A novel approach for improved biomedical data annotation

With the increasing utilization of exome and genome sequencing in clinical and research genetics, accurate and automated extraction of human phenotype ontology (HPO) terms from clinical texts has become imperative. Traditional methods for HPO term extraction, such as PhenoTagger, often face limitations in coverage and precision. In this study, we propose a novel approach that leverages large language models (LLMs) to generate synthetic sentences with clinical context, which were semantically encoded into vector embeddings. These embeddings are linked to HPO terms, creating a robust knowledgebase that facilitates precise information retrieval. Our method circumvents the known issue of LLM hallucinations by storing and querying these embeddings within a true database, ensuring accurate context matching without the need for a predictive model. We evaluated the performance of three different embedding models, all of which demonstrated substantial improvements over PhenoTagger. Top recall (sensitivity), precision (positive-predictive value, PPV), and F1 are 0.64, 0.64, and 0.64, respectively, which were 31%, 10%, and 21% better than PhenoTagger. Furthermore, optimal performance was achieved when we combined the best performing embedding model with PhenoTagger (a.k.a. Fused model), resulting in recall (sensitivity), precision (PPV), and F1 values of 0.7, 0.7, and 0.7, respectively, which are 10%, 10%, and 10% better than the best embedding models. Our findings underscore the potential of this integrated approach to enhance the precision and reliability of HPO term extraction, offering a scalable and effective solution for biomedical data annotation.