Wei Yan

Wei Yan

Auto-generated profile, based on publicly available information

Wei Yan


Research Focus & Metrics

Academic Career Length

15Years (2012-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
24 July 2025 in Sensors
Verified
Journal Article
Open access
PDF file
YZYu Zhou
Yu Zhou
YWYan Wei
Yan Wei
To mitigate the technical challenges associated with small-object detection, feature degradation, and spatial-contextual misalignment in UAV-acquired imagery, this paper proposes UAV-DETR, an enhanced Transformer-based object detection model designed for aerial scenarios. Specifically, UAV imagery often suffers from feature degradation due to low resolution and complex backgrounds and from semantic-spatial misalignment caused by dynamic shooting conditions. This work addresses these challenges by enhancing feature perception, semantic representation, and spatial alignment. Architecturally extending the RT-DETR framework, UAV-DETR incorporates three novel modules: the Channel-Aware Sensing Module (CAS), the Scale-Optimized Enhancement Pyramid Module (SOEP), and the newly designed Context-Spatial Alignment Module (CSAM), which integrates the functionalities of contextual and spatial calibration. These components collaboratively strengthen multi-scale feature extraction, semantic representation, and spatial-contextual alignment. The CAS module refines the backbone to improve multi-scale feature perception, while SOEP enhances semantic richness in shallow layers through lightweight channel-weighted fusion. CSAM further optimizes the hybrid encoder by simultaneously correcting contextual inconsistencies and spatial misalignments during feature fusion, enabling more precise cross-scale integration. Comprehensive comparisons with mainstream detectors, including Faster R-CNN and YOLOv5, demonstrate that UAV-DETR achieves superior small-object detection performance in complex aerial scenarios. The performance is thoroughly evaluated in terms of mAP@0.5, parameter count, and computational complexity (GFLOPs). Experiments on the VisDrone2019 dataset benchmark demonstrate that UAV-DETR achieves an mAP@0.5 of 51.6%, surpassing RT-DETR by 3.5% while reducing the number of model parameters from 19.8 million to 16.8 million.

2 May 2024 in Journal of Medical Internet Research
Verified
Journal Article
Open access
ZGZhenyue Gao
Zhenyue Gao
XLXiaoli Liu
Xiaoli Liu
YKYu Kang
Yu Kang
PHPan Hu
Pan Hu
XZXiu Zhang
Xiu Zhang
ZZZhengbo Zhang
Zhengbo Zhang
Background: Clinical notes contain contextualized information beyond structured data related to patients’ past and current health status. Objective: This study aimed to design a multimodal deep learning approach to improve the evaluation precision of hospital outcomes for heart failure (HF) using admission clinical notes and easily collected tabular data. Methods: Data for the development and validation of the multimodal model were retrospectively derived from 3 open-access US databases, including the Medical Information Mart for Intensive Care III v1.4 (MIMIC-III) and MIMIC-IV v1.0, collected from a teaching hospital from 2001 to 2019, and the eICU Collaborative Research Database v1.2, collected from 208 hospitals from 2014 to 2015. The study cohorts consisted of all patients with critical HF. The clinical notes, including chief complaint, history of present illness, physical examination, medical history, and admission medication, as well as clinical variables recorded in electronic health records, were analyzed. We developed a deep learning mortality prediction model for in-hospital patients, which underwent complete internal, prospective, and external evaluation. The Integrated Gradients and SHapley Additive exPlanations (SHAP) methods were used to analyze the importance of risk factors. Results: The study included 9989 (16.4%) patients in the development set, 2497 (14.1%) patients in the internal validation set, 1896 (18.3%) in the prospective validation set, and 7432 (15%) patients in the external validation set. The area under the receiver operating characteristic curve of the models was 0.838 (95% CI 0.827-0.851), 0.849 (95% CI 0.841-0.856), and 0.767 (95% CI 0.762-0.772), for the internal, prospective, and external validation sets, respectively. The area under the receiver operating characteristic curve of the multimodal model outperformed that of the unimodal models in all test sets, and tabular data contributed to higher discrimination. The medical history and physical examination were more useful than other factors in early assessments. Conclusions: The multimodal deep learning model for combining admission notes and clinical tabular data showed promising efficacy as a potentially novel method in evaluating the risk of mortality in patients with HF, providing more accurate and timely decision support.

12 August 2021 in JMIR mHealth and uHealth
Verified
Journal Article
Open access
HXHaoran Xu
Haoran Xu
WYWei Yan
Wei Yan
KLKe Lan
Ke Lan
CMChenbin Ma
Chenbin Ma
DWDi Wu
Di Wu
ZZZhengbo Zhang
Zhengbo Zhang
Background: With the development and promotion of wearable devices and their mobile health (mHealth) apps, physiological signals have become a research hotspot. However, noise is complex in signals obtained from daily lives, making it difficult to analyze the signals automatically and resulting in a high false alarm rate. At present, screening out the high-quality segments of the signals from huge-volume data with few labels remains a problem. Signal quality assessment (SQA) is essential and is able to advance the valuable information mining of signals. Objective: The aims of this study were to design an SQA algorithm based on the unsupervised isolation forest model to classify the signal quality into 3 grades: good, acceptable, and unacceptable; validate the algorithm on labeled data sets; and apply the algorithm on real-world data to evaluate its efficacy. Methods: Data used in this study were collected by a wearable device (SensEcho) from healthy individuals and patients. The observation windows for electrocardiogram (ECG) and respiratory signals were 10 and 30 seconds, respectively. In the experimental procedure, the unlabeled training set was used to train the models. The validation and test sets were labeled according to preset criteria and used to evaluate the classification performance quantitatively. The validation set consisted of 3460 and 2086 windows of ECG and respiratory signals, respectively, whereas the test set was made up of 4686 and 3341 windows of signals, respectively. The algorithm was also compared with self-organizing maps (SOMs) and 4 classic supervised models (logistic regression, random forest, support vector machine, and extreme gradient boosting). One case validation was illustrated to show the application effect. The algorithm was then applied to 1144 cases of ECG signals collected from patients and the detected arrhythmia false alarms were calculated. Results: The quantitative results showed that the ECG SQA model achieved 94.97% and 95.58% accuracy on the validation and test sets, respectively, whereas the respiratory SQA model achieved 81.06% and 86.20% accuracy on the validation and test sets, respectively. The algorithm was superior to SOM and achieved moderate performance when compared with the supervised models. The example case showed that the algorithm was able to correctly classify the signal quality even when there were complex pathological changes in the signals. The algorithm application results indicated that some specific types of arrhythmia false alarms such as tachycardia, atrial premature beat, and ventricular premature beat could be significantly reduced with the help of the algorithm. Conclusions: This study verified the feasibility of applying the anomaly detection unsupervised model to SQA. The application scenarios include reducing the false alarm rate of the device and selecting signal segments that can be used for further research.

15 April 2021 in JMIR Medical Informatics
Verified
Journal Article
Open access
XLXiaoli Liu
Xiaoli Liu
TLTongbo Liu
Tongbo Liu
ZZZhengbo Zhang
Zhengbo Zhang
PKPo-Chih Kuo
Po-Chih Kuo
HXHaoran Xu
Haoran Xu
DLDeyu Li
Deyu Li
Background: Without timely diagnosis and treatment, tachycardia, also called tachyarrhythmia, can cause serious complications such as heart failure, cardiac arrest, and even death. The predictive performance of conventional clinical diagnostic procedures needs improvement in order to assist physicians in detecting risk early on.Objective: We aimed to develop a deep tachycardia onset prediction (TOP-Net) model based on deep learning (ie, bidirectional long short-term memory) for early tachycardia diagnosis with easily accessible data.Methods: TOP-Net leverages 2 easily accessible data sources: vital signs, including heart rate, respiratory rate, and blood oxygen saturation (SpO2) acquired continuously by wearable embedded systems, and electronic health records, containing age, gender, admission type, first care unit, and cardiovascular disease history. The model was trained with a large data set from an intensive care unit and then transferred to a real-world scenario in the general ward. In this study, 3 experiments incorporated merging patients’ personal information, temporal memory, and different feature combinations. Six metrics (area under the receiver operating characteristic curve [AUROC], sensitivity, specificity, accuracy, F1 score, and precision) were used to evaluate predictive performance.Results: TOP-Net outperformed the baseline models on the large critical care data set (AUROC 0.796, 95% CI 0.768-0.824; sensitivity 0.753, 95% CI 0.663-0.793; specificity 0.720, 95% CI 0.645-0.758; accuracy 0.721; F1 score 0.718; precision 0.686) when predicting tachycardia onset 6 hours in advance. When predicting tachycardia onset 2 hours in advance with data acquired from our hospital using the transferred TOP-Net, the 6 metrics were 0.965, 0.955, 0.881, 0.937, 0.793, and 0.680, respectively. The best performance was achieved using comprehensive vital signs (heart rate, respiratory rate, and SpO2) statistical information.Conclusions: TOP-Net is an early tachycardia prediction model that uses 8 types of data from wearable sensors and electronic health records. When validated in clinical scenarios, the model achieved a prediction performance that outperformed baseline models 0 to 6 hours before tachycardia onset in the intensive care unit and 2 hours before tachycardia onset in the general ward. Because of the model’s implementation and use of easily accessible data from wearable sensors, the model can assist physicians with early discovery of patients at risk in general wards and houses.

12 June 2025 in Remote Sensing
Verified
Journal Article
Open access
PDF file
YLYuling Liu
Yuling Liu
YWYan Wei
Yan Wei
The growth of maritime transportation has been accompanied by a gradual increase in accident rates, drawing greater attention to the critical issue of man-overboard incidents and drowning. Traditional maritime search-and-rescue (SAR) methods are often constrained by limited efficiency and high operational costs. Over the past few years, drones have demonstrated significant promise in improving the effectiveness of search-and-rescue operations. This is largely due to their exceptional ability to move freely and their capacity for wide-area monitoring. This study proposes an enhanced SR-DETR algorithm aimed at improving the detection of individuals who have fallen overboard. Specifically, the conventional multi-head self-attention (MHSA) mechanism is replaced with Efficient Additive Attention (EAA), which facilitates more efficient feature interaction while substantially reducing computational complexity. Moreover, we introduce a new feature aggregation module called the Cross-Stage Partial Parallel Atrous Feature Pyramid Network (CPAFPN). By refining spatial attention mechanisms, the module significantly boosts cross-scale target recognition capabilities in the model, especially offering advantages for detecting smaller objects. To improve localization precision, we develop a novel loss function for bounding box regression, named Focaler-GIoU, which performs particularly well when handling densely packed and small-scale objects. The proposed approach is validated through experiments and achieves an mAP of 86.5%, which surpasses the baseline RT-DETR model’s performance of 83.2%. These outcomes highlight the practicality and reliability of our method in detecting individuals overboard, contributing to more precise and resource-efficient solutions for real-time maritime rescue efforts.

20 June 2025 in IEEE Access
Verified
Journal Article
Open access
ZWZhihao Wei
Zhihao Wei
YWYan Wei
Yan Wei
In modern power systems, insulators, as key components of transmission lines, are crucial for defect detection for the safe operation of power grids. Aiming at the problems of low efficiency of traditional manual detection, the vulnerability of traditional image processing methods to environmental interference, and the insufficient ability of existing deep learning models to detect small target defects under complex backgrounds, this paper proposes an improved target detection model based on YOLOv10n, which is the first time to integrate the spatial channel attention mechanism (SEAttention) with the up-sampling expansion operation (Patch Expanding) in the Neck part of the model. In the Neck part of the model, the spatial channel attention mechanism (SEAttention) and the up-sampling expansion operation (Patch Expanding) are deeply integrated. The spatial information of the channels is compressed by global average pooling, and the feature map is adaptively weighted after learning the channel weights through the fully connected layer, so as to strengthen the key channel features and suppress the noise. The Concat module in Neck is replaced by a two-way weighted feature fusion mechanism, which constructs a top-down and bottom-up two-way information flow, reduces the loss of cross-layer information, and significantly enhances the model’s ability to detect multi-scale defects, such as insulator cracks and fouling, and especially improves the resolution of details of small targets in long-distance shooting scenarios. The channel-space dual attention mechanism (CBAM) is integrated into the C2f module, and the channel weights are computed by global average pooling and maximum pooling in parallel, combined with the spatial attention features extracted by convolution, to realize the accurate focusing on defect-related channels and spatial regions, and the experimental results show that the improved model’s mAP@50 reaches 94.2%, which is 2.4% better than that of the baseline model YOLOv10n, and better than that of YOLOv5n. The experimental results show that the mAP@50 of the improved model reaches 94.2%, which is 2.4% higher than that of the baseline model YOLOv10n, and better than the mainstream algorithms such as YOLOv5 (93.1%) and YOLOv8n (93.2%).