http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
환경변화에 강인한 딥러닝 기반의 터널 균열 측정 및 진단
L. Minh Dang,Chanmi Oh,Yanfen Li,Hanxiang Wang,Chang-Jae Chun,Hyeonjoon Moon 한국차세대컴퓨팅학회 2022 한국차세대컴퓨팅학회 학술대회 Vol.2022 No.05
A tunnel is an essential public facility that enables uninterrupted transportation in crowded cities. Over time, various factors such as ageing and harsh environment could slowly damage the tunnel, leading to cracks and even human loss. There, the tunnel needs to be investigated regularly. Previous maintenance methods have primarily counted on the operators who directly monitor recorded videos to inspect the cracks and determine their seriousness. However, this is a time-consuming and error-prone process. Firstly, this paper introduces a huge tunnel cracks segmentation dataset that contains a total of 170,339 images. Next, a tunnel crack segmentation system that can automatically identify different types of cracks is suggested based on the collected data. The model uses the U-Net structure as the baseline model, with the encoder replaced by a pre-trained Resnet-152 model to improve the effectiveness of the feature extract process. Finally, additional measurements of the detected cracks, such as crack length and crack thickness, are computed.
Robust masonry crack segmentation and measurement based on deep learning
L. Minh Dang,Le Quan Nguyen,Shin Jihye,Hyeonjoon Moon 한국차세대컴퓨팅학회 2022 한국차세대컴퓨팅학회 학술대회 Vol.2022 No.10
Masonry is a common type of construction that uses mortar to bind individual units, such as brick or building stones, together to construct the structure. Even though masonry structures are durable, multiple factors such as the quality of mortar, workmanship, and harsh environment could greatly reduce the structural integrity, leading to defects and even human loss. Thus, it is crucial to perform the maintenance process regularly. Previously, the maintenance relied mainly on inspectors, who inspected the masonry structures to find cracks and determine the seriousness. However, this process is error-prone, costly, and time-consuming. As a result, this study proposes a fully automated masonry crack segmentation framework that robustly identifies various types of masonry cracks. In addition, the length of the segmentation cracks, which has been ignored in previous studies, is also computed.
Tran Dang Thanh,Phan, T. L.,Le Mai Oanh,Nguyen Van Minh,Jong Suk Lee,Yu, S. C. IEEE 2014 IEEE transactions on magnetics Vol.50 No.6
<P>This paper presents the influence of Mn doping on the structural characterization, and optical and magnetic properties of SrTi<SUB>1-x</SUB>Mn<SUB>x</SUB>O<SUB>3</SUB>(x = 0.0-0.1) materials prepared by a solid-state reaction method. The detailed analyses of X-ray diffraction patterns indicate an incorporation of Mn dopants into Ti sites of the SrTiO<SUB>3</SUB> host lattice. There is a cubic to tetragonal transformation, which takes place at a threshold concentration x ≈ 0.04. The optical absorption spectra show a rapid increase in the absorption coefficient. The bandgap energy (Eg) related to the direct electron transition decreases with increasing Mn concentration: Eg decreases from 3.15 eV for x = 0 to 1.28 eV for x = 0.10. From this point of view, the SrTi<SUB>1-x</SUB>Mn<SUB>x</SUB>O<SUB>3</SUB> materials are considered as promising materials for photocatalytic applications. Interestingly, while the samples with x = 0.0-0.02 are diamagnetic, the others with x = 0.04-0.10 exhibit weak ferromagnetism. The ferromagnetic order increases with increasing Mn concentration. Based on the results of structural and optical analyses, the nature of magnetism in the samples is explained thoroughly.</P>
Wenqi Zhang,L. Minh Dang,Yanfen Li,Hanxiang Wang,Sujin Lee,Hyeonjoon Moon 한국방송·미디어공학회 2023 한국방송공학회 학술발표대회 논문집 Vol.2023 No.6
The dimensions of plants, including their length and width, as well as the size of their leaves, serve as crucial indicators for assessing their growth status. These factors provide essential data for studying plant conditions. This article introduces an algorithm for instance segmentation based on Solo, with the addition of a residual block module to enhance segmentation performance. The original rectified linear unit (ReLU) activation function is replaced with a functional ReLU (PReLU), and an open turnip segmentation dataset is proposed. Experimental results demonstrate that, in comparison to the original model, the average accuracy of the modified model reaches 87.6%, an improvement of approximately 2.0%. The improved algorithm accurately segments turnips and their leaves, exhibiting superior accuracy in measuring both turnip and leaf dimensions. Compared to manual measurements, the average accuracy for turnip length is 97.38%, turnip width is 95.46%, turnip leaf length is 97.79%, and turnip leaf width is 96.13%.
민경복 ( Kyungbok Min ),( L. Minh Dang ),이수진 ( Sujin Lee ),문현준 ( Hyeonjoon Moon ) 한국정보처리학회 2019 한국정보처리학회 학술대회논문집 Vol.26 No.2
Research using artificial intelligence to generate captions for an image has been studied extensively. However, these systems are unable to create creative stories that include more than one sentence based on image content. A story is a better way that humans use to foster social cooperation and develop social norms. This paper proposes a framework that can generate a relatively short story to describe based on the context of an image. The main contributions of this paper are (1) An unsupervised framework which uses recurrent neural network structure and encoder-decoder model to construct a short story for an image. (2) A huge English novel dataset, including horror and romantic themes that are manually collected and validated. By investigating the short stories, the proposed model proves that it can generate more creative contents compared to existing intelligent systems which can produce only one concise sentence. Therefore, the framework demonstrated in this work will trigger the research of a more robust AI story writer and encourages the application of the proposed model in helping story writer find a new idea.
Muhammad Nadeem,Haseeb Khan,Wisal Khan,L. Minh Dang,Nguyen Le Quan,Hyeonjoon Moon 한국차세대컴퓨팅학회 2023 한국차세대컴퓨팅학회 학술대회 Vol.2023 No.12
Breast cancer remains the foremost cause of cancer-related mortality worldwide. The histopathological diagnosis is impeded by the intricate nature of image interpretation and the presence of inter-observer variability among pathologists. Deep learning (DL) for cancer image understanding has revolutionized accurate breast cancer diagnosis, marking a significant advancement in medical image analysis. Researchers proposed DL-based intelligent models to overcome the challenges of manual observations. However, the existing models suffer from a considerable computational burden, demanding substantial time investments that restrict efficient and scalable breast cancer diagnosis solutions. Our study introduces an automated breast cancer diagnosis system employing a lightweight Convolutional Neural Network (CNN) model, adept at extracting intricate features from histopathological images. Our system has attained superior accuracy through extensive experimentation on a comprehensive breast cancer dataset while employing fewer parameters compared to state-of-theart (SOTA) techniques.