• Journal of the Korean Geotechnical Society
• /
• v.40 no.3
• /
• pp.33-39
• /
• 2024

Inspecting cracks to determine a structure's condition is crucial for accurate safety diagnosis. However, visual crack inspection methods can be subjective and are dependent on field conditions, thereby resulting in low reliability. To address this issue, this study automates the detection of concrete cracks in image data using ResNet, FPN, and the Mask R-CNN components as the backbone, neck, and head of a convolutional neural network. The performance of the proposed model is analyzed using the intersection over the union (IoU). The experimental dataset contained 1,203 images divided into training (70%), validation (20%), and testing (10%) sets. The model achieved an IoU value of 95.83% for testing, and there were no cases where the crack was not detected. These findings demonstrate that the proposed model realized highly accurate detection of concrete cracks in image data.

• Journal of Korean Neurosurgical Society
• /
• v.63 no.3
• /
• pp.386-396
• /
• 2020

Objective : To generate synthetic spine magnetic resonance (MR) images from spine computed tomography (CT) using generative adversarial networks (GANs), as well as to determine the similarities between synthesized and real MR images. Methods : GANs were trained to transform spine CT image slices into spine magnetic resonance T2 weighted (MRT2) axial image slices by combining adversarial loss and voxel-wise loss. Experiments were performed using 280 pairs of lumbar spine CT scans and MRT2 images. The MRT2 images were then synthesized from 15 other spine CT scans. To evaluate whether the synthetic MR images were realistic, two radiologists, two spine surgeons, and two residents blindly classified the real and synthetic MRT2 images. Two experienced radiologists then evaluated the similarities between subdivisions of the real and synthetic MRT2 images. Quantitative analysis of the synthetic MRT2 images was performed using the mean absolute error (MAE) and peak signal-to-noise ratio (PSNR). Results : The mean overall similarity of the synthetic MRT2 images evaluated by radiologists was 80.2%. In the blind classification of the real MRT2 images, the failure rate ranged from 0% to 40%. The MAE value of each image ranged from 13.75 to 34.24 pixels (mean, 21.19 pixels), and the PSNR of each image ranged from 61.96 to 68.16 dB (mean, 64.92 dB). Conclusion : This was the first study to apply GANs to synthesize spine MR images from CT images. Despite the small dataset of 280 pairs, the synthetic MR images were relatively well implemented. Synthesis of medical images using GANs is a new paradigm of artificial intelligence application in medical imaging. We expect that synthesis of MR images from spine CT images using GANs will improve the diagnostic usefulness of CT. To better inform the clinical applications of this technique, further studies are needed involving a large dataset, a variety of pathologies, and other MR sequence of the lumbar spine.

• The Korean Journal of Applied Statistics
• /
• v.36 no.2
• /
• pp.101-114
• /
• 2023

The data augmentation technique is effectively used to solve the problem of overfitting the model by allowing the training dataset to be viewed from various perspectives. In addition to image augmentation techniques such as rotation, cropping, horizontal flip, and vertical flip, occlusion-based data augmentation methods such as Cutmix and Cutout have been proposed. For models based on speech data, it is possible to use an occlusion-based data-based augmentation technique after converting a 1D speech signal into a 2D spectrogram. In particular, SpecAugment is an occlusion-based augmentation technique for speech spectrograms. In this study, we intend to compare and study data augmentation techniques that can be used in the problem of false-voice detection. Using data from the ASVspoof2017 and ASVspoof2019 competitions held to detect fake audio, a dataset applied with Cutout, Cutmix, and SpecAugment, an occlusion-based data augmentation method, was trained through an LCNN model. All three augmentation techniques, Cutout, Cutmix, and SpecAugment, generally improved the performance of the model. In ASVspoof2017, Cutmix, in ASVspoof2019 LA, Mixup, and in ASVspoof2019 PA, SpecAugment showed the best performance. In addition, increasing the number of masks for SpecAugment helps to improve performance. In conclusion, it is understood that the appropriate augmentation technique differs depending on the situation and data.

• Journal of Intelligence and Information Systems
• /
• v.29 no.2
• /
• pp.323-338
• /
• 2023

TV advertising with deep analysis of watching pattern of audiences is important to set-top box audience targeting. Applying session-based recommendation model(SBR) to internet commercial, or recommendation based on searching history of user showed its effectiveness in previous studies, but applying SBR to the TV advertising was difficult in South Korea due to data unavailabilities. Also, traditional SBR has limitations for dealing with user preferences, especially in data with user identification information. To tackle with these problems, we first obtain set-top box data from three major broadcasting companies in South Korea(SKB, KT, LGU+) through collaboration with Korea Broadcast Advertising Corporation(KOBACO), and this data contains of watching sequence of 4,847 anonymized users for 6 month respectively. Second, we develop personalized session-based recommendation model to deal with hierarchical data of user-session-item. Experiments conducted on set-top box audience dataset and two other public dataset for validation. In result, our proposed model outperformed baseline model in some criteria.

• Journal of The Korean Association For Science Education
• /
• v.43 no.3
• /
• pp.237-251
• /
• 2023

This study explores the possibility of automated scoring for scientific graph answers by designing an automated scoring model using convolutional neural networks and applying it to students' kinematics graph answers. The researchers prepared 2,200 answers, which were divided into 2,000 training data and 200 validation data. Additionally, 202 student answers were divided into 100 training data and 102 test data. First, in the process of designing an automated scoring model and validating its performance, the automated scoring model was optimized for graph image classification using the answer dataset prepared by the researchers. Next, the automated scoring model was trained using various types of training datasets, and it was used to score the student test dataset. The performance of the automated scoring model has been improved as the amount of training data increased in amount and diversity. Finally, compared to human scoring, the accuracy was 97.06%, the kappa coefficient was 0.957, and the weighted kappa coefficient was 0.968. On the other hand, in the case of answer types that were not included in the training data, the s coring was almos t identical among human s corers however, the automated scoring model performed inaccurately.

• Journal of IKEEE
• /
• v.27 no.4
• /
• pp.464-471
• /
• 2023

In this paper, we propose a method to reduce age distortion in facial expression image generation using StyleGAN Encoder. The facial expression image generation process first creates a face image using StyleGAN Encoder, and changes the expression by applying the learned boundary to the latent vector using SVM. However, when learning the boundary of a smiling expression, age distortion occurs due to changes in facial expression. The smile boundary created in SVM learning for smiling expressions includes wrinkles caused by changes in facial expressions as learning elements, and it is determined that age characteristics were also learned. To solve this problem, the proposed method calculates the correlation coefficient between the smile boundary and the age boundary and uses this to introduce a method of adjusting the age boundary at the smile boundary in proportion to the correlation coefficient. To confirm the effectiveness of the proposed method, the results of an experiment using the FFHQ dataset, a publicly available standard face dataset, and measuring the FID score are as follows. In the smile image, compared to the existing method, the FID score of the smile image generated by the ground truth and the proposed method was improved by about 0.46. In addition, compared to the existing method in the smile image, the FID score of the image generated by StyleGAN Encoder and the smile image generated by the proposed method improved by about 1.031. In non-smile images, compared to the existing method, the FID score of the non-smile image generated by the ground truth and the method proposed in this paper was improved by about 2.25. In addition, compared to the existing method in non-smile images, it was confirmed that the FID score of the image generated by StyleGAN Encoder and the non-smile image generated by the proposed method improved by about 1.908. Meanwhile, as a result of estimating the age of each generated facial expression image and measuring the estimated age and MSE of the image generated with StyleGAN Encoder, compared to the existing method, the proposed method has an average age of about 1.5 in smile images and about 1.63 in non-smile images. Performance was improved, proving the effectiveness of the proposed method.

• The Journal of the Acoustical Society of Korea
• /
• v.43 no.1
• /
• pp.9-18
• /
• 2024

The importance of active sonar systems is emerging due to the quietness of underwater targets and the increase in ambient noise due to the increase in maritime traffic. However, the low signal-to-noise ratio of the echo signal due to multipath propagation of the signal, various clutter, ambient noise and reverberation makes it difficult to identify underwater targets using active sonar. Attempts have been made to apply data-based methods such as machine learning or deep learning to improve the performance of underwater target recognition systems, but it is difficult to collect enough data for training due to the nature of sonar datasets. Methods based on mathematical modeling have been mainly used to compensate for insufficient active sonar data. However, methodologies based on mathematical modeling have limitations in accurately simulating complex underwater phenomena. Therefore, in this paper, we propose a sonar signal synthesis method based on a deep neural network. In order to apply the neural network model to the field of sonar signal synthesis, the proposed method appropriately corrects the attention-based encoder and decoder to the sonar signal, which is the main module of the Tacotron model mainly used in the field of speech synthesis. It is possible to synthesize a signal more similar to the actual signal by training the proposed model using the dataset collected by arranging a simulated target in an actual marine environment. In order to verify the performance of the proposed method, Perceptual evaluation of audio quality test was conducted and within score difference -2.3 was shown compared to actual signal in a total of four different environments. These results prove that the active sonar signal generated by the proposed method approximates the actual signal.

• Tunnel and Underground Space
• /
• v.32 no.2
• /
• pp.131-143
• /
• 2022

The lunar exploration autonomous vehicle operates based on the lunar topography information obtained from real-time image characterization. For highly accurate topography characterization, a large number of training images with various background conditions are required. Since the real lunar topography images are difficult to obtain, it should be helpful to be able to generate mimic lunar image data artificially on the basis of the planetary analogs site images and real lunar images available. In this study, we aim to artificially create lunar topography images by using the location information-based style transfer algorithm known as Wavelet Correct Transform (WCT2). We conducted comparative experiments using lunar analog site images and real lunar topography images taken during China's and America's lunar-exploring projects (i.e., Chang'e and Apollo) to assess the efficacy of our suggested approach. The results show that the proposed techniques can create realistic images, which preserve the topography information of the analog site image while still showing the same condition as an image taken on lunar surface. The proposed algorithm also outperforms a conventional algorithm, Deep Photo Style Transfer (DPST) in terms of temporal and visual aspects. For future work, we intend to use the generated styled image data in combination with real image data for training lunar topography objects to be applied for topographic detection and segmentation. It is expected that this approach can significantly improve the performance of detection and segmentation models on real lunar topography images.

• Journal of radiological science and technology
• /
• v.46 no.4
• /
• pp.315-323
• /
• 2023

Post-mortem computed tomography (PMCT) is commonly employed in the field of forensic pathology. PMCT was mainly performed using a whole-body scan with a wide field of view (FOV), which lead to a decrease in spatial resolution due to the increased pixel size. This study aims to evaluate the potential for developing a super-resolution model based on conditional generative adversarial networks (CGAN) to enhance the image quality of CT. 1761 low-resolution images were obtained using a whole-body scan with a wide FOV of the head phantom, and 341 high-resolution images were obtained using the appropriate FOV for the head phantom. Of the 150 paired images in the total dataset, which were divided into training set (96 paired images) and validation set (54 paired images). Data augmentation was perform to improve the effectiveness of training by implementing rotations and flips. To evaluate the performance of the proposed model, we used the Peak Signal-to-Noise Ratio (PSNR), Structural Similarity Index Measure (SSIM) and Deep Image Structure and Texture Similarity (DISTS). Obtained the PSNR, SSIM, and DISTS values of the entire image and the Medial orbital wall, the zygomatic arch, and the temporal bone, where fractures often occur during head trauma. The proposed method demonstrated improvements in values of PSNR by 13.14%, SSIM by 13.10% and DISTS by 45.45% when compared to low-resolution images. The image quality of the three areas where fractures commonly occur during head trauma has also improved compared to low-resolution images.

• The Journal of the Acoustical Society of Korea
• /
• v.39 no.5
• /
• pp.414-423
• /
• 2020

In this paper, we propose a Dilated Convolution Gate Linear Unit (DCGLU) to mitigate the lack of sparsity and small receptive field problems caused by the segmentation map extraction process in sound event detection with weak labels. In the advent of deep learning framework, segmentation map extraction approaches have shown improved performance in noisy environments. However, these methods are forced to maintain the size of the feature map to extract the segmentation map as the model would be constructed without a pooling operation. As a result, the performance of these methods is deteriorated with a lack of sparsity and a small receptive field. To mitigate these problems, we utilize GLU to control the flow of information and Dilated Convolutional Neural Networks (DCNNs) to increase the receptive field without additional learning parameters. For the performance evaluation, we employ a URBAN-SED and self-organized bird sound dataset. The relevant experiments show that our proposed DCGLU model outperforms over other baselines. In particular, our method is shown to exhibit robustness against nature sound noises with three Signal to Noise Ratio (SNR) levels (20 dB, 10 dB and 0 dB).