ATNet: A Defect Detection Framework for X-ray Images of DIP Chip Lead Bonding

Huang, Renbin; Zhan, Daohua; Yang, Xiuding; Zhou, Bei; Tang, Luping; Cai, Nian; Wang, Han; Qiu, Baojun

doi:10.3390/mi14071375

Cited by 4 publications

(2 citation statements)

References 42 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Looking ahead, large language models such as ChatGPT have the potential to significantly impact the field of healthcare [109], including HF. Additionally, advances in image segmentation technology may bring about new changes in medical image processing [110,111].…”

Section: Discussionmentioning

confidence: 99%

Machine Learning in Heart Failure Research: A Bibliometric Analysis from 2003 to 2023

Yang

Wang

et al. 2023

Preprint

View full text Add to dashboard Cite

Background Heart failure (HF) is a global public health problem with high morbidity and mortality. While machine learning (ML) has been perceived as a promising tool for HF research, a bibliometric analysis of this application is still lacking. This study aims to analyze the relevant papers from 2003 to 2023 and provide a comprehensive overview of this field in a visual way. Methods We systematically searched Web of Science Core Collection, PubMed, and preprint servers (arXiv, BioRxiv, and MedRxiv) to identify records from 2003 to 2023 on ML in HF research. After manual data cleansing, a Python approach based on the regular expression matching algorithm was used to automate text annotations, while three bibliometric analysis tools (CiteSpace, VOSviewer, and Bibliometrix) were used for the visualization of research trends, collaboration networks, and research hotspots. Results We analyzed 6,115 records (including 1,797 published papers) and observed a steady increase in annual publication rates since 2015, with a significant uptick after 2020. We identified 23 core journals in the field according to Bradford's law, and presented the top 10 journals with the highest citations, h-index, g-index, or m-index. The United States was the most productive country, followed by China and the United Kingdom. The most prolific institutions were Harvard Medical School and Mayo Clinic. Using text annotations, we identified 1,257 ML-related original HF research. In these studies, the main data modalities were ultrasound, electronic health records, and electrocardiograms. The most frequently applied ML methods were neural networks, followed by linear models and ensembles. The most common clinical goals were diagnosis, prediction, and classification. The main research topics included the classification of HF, AI-assisted medical diagnostic technologies, HF-related clinical prediction models, and HF-related bioinformatics research. Conclusions This bibliometric analysis revealed a significant growth in the application of ML in HF research over the past two decades. The current research landscape encompasses a wide range of ML techniques and applications, focusing on improving diagnosis, prognosis, classification, and precision treatment for HF patients.

show abstract

Section: Discussionmentioning

confidence: 99%

Machine Learning in Heart Failure Research: A Bibliometric Analysis from 2003 to 2023

Yang

Wang

et al. 2023

Preprint

View full text Add to dashboard Cite

show abstract

“…In the comparative experiments, RPDNet was exhaustively benchmarked against two-stage detection method Faster-RCNN, and one-stage methods RetinaNet [30], YOLOv4-tiny, YOLOv5m, YOLO-x-ray [31], and ATNet [32]. All models were trained on the same dataset and evaluated in a uniform testing environment.…”

Section: Comparative Experimentsmentioning

confidence: 99%

Soldered chip reversed polarity detection based on semi-combinatorial attention and size-sensitive IoU

Zheng,

Wu,

Chen

2024

Meas. Sci. Technol.

View full text Add to dashboard Cite

Chips that have been automatically soldered may have defects such as short circuits, reversed polarity, oﬀsets, and missing components. This paper proposes a reversed polarity detection algorithm (RPDNet) for soldered chips based on semi-combinatorial attention and size-sensitive IoU to solve the polarity reversal problem. To distinguish characters or patterns similar to polarity markings in the background, a semi-combination attention module (SCAM) is designed, which eﬀectively distinguishes the subtle feature diﬀerences between polarity markings and background elements by integrating the long-range dependence of the self-attention mechanism and the local focusing ability of the general attention mechanism, thereby reducing misdetection of polarity markings. Aiming at the problem of blurred chip polarity markings, a differential pooling module (DPM) is proposed, which optimizes the downsampling process of fuzzy polarity markings by combining the differential features obtained from subtracting the results of max pooling from average pooling with features derived solely from average pooling, thereby reducing the information loss of fuzzy markings during the feature extraction process. In order to solve the lack of sensitivity of CIoU to the side length of the bounding box, a size-sensitive IoU(SIoU) is proposed. Through the size-sensitivity factor and area diﬀerence loss term, the model strengthens the focus on boundary regression and achieves faster and more accurate target positioning. Using the Chip-Set provided by the company for experiments and tests, the detection accuracy reached 97.79% and the processing speed reached 107 FPS.

show abstract

Deep learning in person re-identification: a survey

Xiao,

Zhang,

Ning

et al. 2024

International Conference on Image Processing and Artificial Intelligence (ICIPAl 2024)

View full text Add to dashboard Cite

In the era of advancing deep learning, person re-identification has gained widespread application in domains such as video and security monitoring. Person re-identification seeks to recognize and match target persons across images taken by diverse cameras, thereby verifying whether the pedestrian subjects observed by cameras at various locations and times are indeed the same individual. This article categorizes current research into two main types: image-based person reidentification and video-based person re-identification. It provides a thorough overview and analysis of existing literature concerning classification methods, verification models, attention mechanisms, and metric learning,all centered around the research subjects. Additionally, it outlines the progression of datasets and provides insights into the anticipated future trends in person re-identification

show abstract

ATNet: A Defect Detection Framework for X-ray Images of DIP Chip Lead Bonding

Cited by 4 publications

References 42 publications

Machine Learning in Heart Failure Research: A Bibliometric Analysis from 2003 to 2023

Machine Learning in Heart Failure Research: A Bibliometric Analysis from 2003 to 2023

Soldered chip reversed polarity detection based on semi-combinatorial attention and size-sensitive IoU

Deep learning in person re-identification: a survey

Contact Info

Product

Resources

About