• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.15 no.3
• /
• pp.911-931
• /
• 2021

Recently, most cloud services use Docker container environment to provide their services. However, there are no researches to evaluate the performance of communication libraries for multi-GPU based distributed deep learning in a Docker container environment. In this paper, we propose an efficient communication architecture for multi-GPU based deep learning in a Docker container environment by evaluating the performances of various communication libraries. We compare the performances of the parameter server architecture and the All-reduce architecture, which are typical distributed deep learning architectures. Further, we analyze the performances of two separate multi-GPU resource allocation policies - allocating a single GPU to each Docker container and allocating multiple GPUs to each Docker container. We also experiment with the scalability of collective communication by increasing the number of GPUs from one to four. Through experiments, we compare OpenMPI and MPICH, which are representative open source MPI libraries, and NCCL, which is NVIDIA's collective communication library for the multi-GPU setting. In the parameter server architecture, we show that using CUDA-aware OpenMPI with multi-GPU per Docker container environment reduces communication latency by up to 75%. Also, we show that using NCCL in All-reduce architecture reduces communication latency by up to 93% compared to other libraries.

• Smart Structures and Systems
• /
• v.29 no.6
• /
• pp.777-783
• /
• 2022

Even though several deep learning-based methods have been applied in the field of acoustic source localization, the previous works have only been conducted using the two-dimensional representation of the beamforming maps, particularly with the planar array system. While the acoustic sources are more required to be localized in a spherical microphone array system considering that we live and hear in the 3D world, the conventional 2D equirectangular map of the spherical beamforming map is highly vulnerable to the distortion that occurs when the 3D map is projected to the 2D space. In this study, a 3D deep learning approach is proposed to fulfill accurate source localization via distortion-free 3D representation. A target function is first proposed to obtain 3D source distribution maps that can represent multiple sources' positional and strength information. While the proposed target map expands the source localization task into a point-wise prediction task, a PointNet-based deep neural network is developed to precisely estimate the multiple sources' positions and strength information. While the proposed model's localization performance is evaluated, it is shown that the proposed method can achieve improved localization results from both quantitative and qualitative perspectives.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.18 no.1
• /
• pp.99-106
• /
• 2023

Recently, many studies on crop yield prediction using deep learning technology have been conducted. These algorithms have difficulty constructing a linear map between input data sets and crop prediction results. Furthermore, implementation of these algorithms positively depends on the rate of acquired attributes. Deep reinforcement learning can overcome these limitations. This paper analyzes the performance of DQN, Double DQN and Dueling DQN to improve crop yield prediction. The DQN algorithm retains the overestimation problem. Whereas, Double DQN declines the over-estimations and leads to getting better results. The proposed models achieves these by reducing the falsehood and increasing the prediction exactness.

• Language Facts and Perspectives
• /
• v.48
• /
• pp.195-227
• /
• 2019

Building upon the state-of-the-art deep learning techniques, the present study classifies the texts written by Korean EFL learners and English native speakers and thereby demonstrates how the two types of texts differ from each other. To this end, the current work makes use of the Yonsei English Learner Corpus (YELC) and Gacheon Learner Corpus (GLC) as the L2 data, and Corpus of Contemporary American English (COCA) as the L1 data. Utilizing the sentence classification methods, the current work implements a system to differentiate the two types of texts, the accuracy of which is about 94%. This indicates that the deep leaning-based system is capable of identifying the well-formedness and felicities of the texts written by Korean EFL learners. Nonetheless, the system-based judgments do not overlap with human judgments largely because the deep learning model exclusively focuses on sequence of words. The present study provides a further analysis to see how the two types of judgments differ with respect to grammatical errors (e.g., word order, voice, etc.) and felicity errors (e.g., semantic prosody, the position of adverbs, etc.).

• Journal of The Korean Association of Information Education
• /
• v.23 no.6
• /
• pp.583-590
• /
• 2019

The purpose of this study is to develop an educational program for learning deep learning concepts for elementary school students. The model of education program was developed the deep-learning teaching method based on CT element-oriented teaching and learning model. The subject of the developed program is the artificial intelligence image recognition CNN algorithm, and we have developed 9 educational programs. We applied the program over two weeks to sixth graders. Expert validity analysis showed that the minimum CVR value was more than .56. The fitness level of learner level and the level of teacher guidance were less than .80, and the fitness of learning environment and media above .96 was high. The students' satisfaction analysis showed that students gave a positive evaluation of the average of 4.0 or higher on the understanding, benefit, interest, and learning materials of artificial intelligence learning.

• Journal of the Korean Institute of Gas
• /
• v.27 no.2
• /
• pp.49-56
• /
• 2023

As AI (Artificial Intelligence)-related technologies are highly developed due to the 4th industrial revolution, cases of applying AI in various fields are increasing. The main reason is that there are practical limits to direct processing and analysis of exponentially increasing data as information and communication technology develops, and the risk of human error can be reduced by applying new technologies. In this study, after collecting the data received from the 'remote potential measurement terminal (T/B, Test Box)' and the output of the 'remote rectifier' at that time, AI was trained. AI learning data was obtained through data augmentation through regression analysis of the initially collected data, and the learning model applied the value-based Q-Learning model among deep reinforcement learning (DRL) algorithms. did The AI that has completed data learning is put into the actual city gas supply area, and based on the received remote T/B data, it is verified that the AI responds appropriately, and through this, AI can be used as a suitable means for electricity management in the future. want to verify.

• Composites Research
• /
• v.36 no.2
• /
• pp.92-100
• /
• 2023

In this paper, a machine learning and deep learning model for the design of circuit analog (CA) radar absorbing structure with a cross-dipole pattern on a glass fiber fabric reinforced plastic is presented. The proposed model can directly calculate reflection loss in the Ku-band (12-18 GHz) without three-dimensional electromagnetic numerical analysis based on the geometry of the Cross-Dipole pattern. For this purpose, the optimal learning model was derived by applying various machine learning and deep learning techniques, and the results calculated by the learning model were compared with the electromagnetic wave absorption characteristics obtained by 3D electromagnetic wave numerical analysis to evaluate the comparative advantages of each model. Most of the implemented models showed similar calculated results to the numerical results, but it was found that the Fully-Connected model could provide the most similar calculated results.

• The Journal of the Acoustical Society of Korea
• /
• v.40 no.2
• /
• pp.169-175
• /
• 2021

Recently, an ultrasound hologram and its applications have gained attention in the ultrasound research field. However, the determination technique of transmit signal phases, which generate a hologram, has not been significantly advanced from the previous algorithms which are time-consuming iterative methods. Thus, we applied the deep learning technique, which has been previously adopted to generate an optical hologram, to generate an ultrasound hologram. We further examined the Deep learning-based Holographic Ultrasound Generation algorithm (Deep-HUG). We implement the U-Net-based algorithm and examine its generalizability by training on a dataset, which consists of randomly distributed disks, and testing on the alphabets (A-Z). Furthermore, we compare the Deep-HUG with the previous algorithm in terms of computation time, accuracy, and uniformity. It was found that the accuracy and uniformity of the Deep-HUG are somewhat lower than those of the previous algorithm whereas the computation time is 190 times faster than that of the previous algorithm, demonstrating that Deep-HUG has potential as a useful technique to rapidly generate an ultrasound hologram for various applications.

• Korean Journal of Remote Sensing
• /
• v.38 no.6_1
• /
• pp.1091-1100
• /
• 2022

As deep learning technology advances and various high-resolution remote sensing images are available, interest in using deep learning technology and remote sensing big data to detect buildings and change in urban areas is increasing significantly. In this paper, for semantic building segmentation of high-resolution remote sensing images, we propose a new building segmentation model, Convolutional Block Attention Module (CBAM)-DRUNet that uses the DeepResUNet model, which has excellent performance in building segmentation, as the basic structure, improves the residual learning unit and combines a CBAM with the basic structure. In the performance evaluation using WHU dataset and INRIA dataset, the proposed building segmentation model showed excellent performance in terms of F1 score, accuracy and recall compared to ResUNet and DeepResUNet including UNet.

• Structural Engineering and Mechanics
• /
• v.91 no.5
• /
• pp.443-457
• /
• 2024

At present, the traditional concrete surface inspection methods based on artificial vision have the problems of high cost and insecurity, while the computer vision methods rely on artificial selection features in the case of sensitive environmental changes and difficult promotion. In order to solve these problems, this paper introduces deep learning technology in the field of computer vision to achieve automatic feature extraction of structural damage, with excellent detection speed and strong generalization ability. The main contents of this study are as follows: (1) A method based on DeepLabV3+ convolutional neural network model is proposed for surface detection of post-earthquake structural damage, including surface damage such as concrete cracks, spaling and exposed steel bars. The key semantic information is extracted by different backbone networks, and the data sets containing various surface damage are trained, tested and evaluated. The intersection ratios of 54.4%, 44.2%, and 89.9% in the test set demonstrate the network's capability to accurately identify different types of structural surface damages in pixel-level segmentation, highlighting its effectiveness in varied testing scenarios. (2) A semantic segmentation model based on DeepLabV3+ convolutional neural network is proposed for the detection and evaluation of post-earthquake structural components. Using a dataset that includes building structural components and their damage degrees for training, testing, and evaluation, semantic segmentation detection accuracies were recorded at 98.5% and 56.9%. To provide a comprehensive assessment that considers both false positives and false negatives, the Mean Intersection over Union (Mean IoU) was employed as the primary evaluation metric. This choice ensures that the network's performance in detecting and evaluating pixel-level damage in post-earthquake structural components is evaluated uniformly across all experiments. By incorporating deep learning technology, this study not only offers an innovative solution for accurately identifying post-earthquake damage in civil engineering structures but also contributes significantly to empirical research in automated detection and evaluation within the field of structural health monitoring.