• 한국지반환경공학회 논문집
• /
• 제22권10호
• /
• pp.13-19
• /
• 2021

아스팔트 포장의 균열은 날씨의 변화나 차량에 의한 충격으로 발생하며, 균열을 방치할 경우 포장 수명이 단축되고 각종 사고를 불러 일으킬 수 있다. 따라서 아스팔트 도로 포장의 균열을 빠르게 감지하여 보수조치를 취하기 위하여 이미지를 통해 균열을 자동으로 탐지하기 위한 연구들이 지속되어 왔다. 특히 최근들어 Convolutional Neural Network를 사용하여 아스팔트 도로 포장의 균열을 탐지하려는 모델들이 많이 연구되고 있으나, 고성능의 컴퓨팅 파워를 요구하기 때문에 실제 활용에는 한계가 있다. 이에 본 논문에서는 모바일 기기에 적용 가능한 스몰 딥러닝 모델을 적용하여 아스팔트 도로 포장의 균열을 탐지하는 모델의 개발을 위한 프레임워크를 제안한다. 사례연구를 통해 제안한 스몰 딥러닝 모델은 일반적인 딥러닝 모델들과 비교 연구되었으며, 상대적으로 적은 파라미터를 가지는 모델임에도 일반적인 딥러닝 모델들과 유사한 성능을 보였다. 개발된 모델은 모바일 기기나 IoT에 임베디드 되어 사용될 수 있을 것으로 기대된다.

• 한국재료학회지
• /
• 제32권8호
• /
• pp.345-353
• /
• 2022

In this study, using deep learning, super-resolution images of transmission electron microscope (TEM) images were generated for nanomaterial analysis. 1169 paired images with 256 × 256 pixels (high resolution: HR) from TEM measurements and 32 × 32 pixels (low resolution: LR) produced using the python module openCV were trained with deep learning models. The TEM images were related to DyVO₄ nanomaterials synthesized by hydrothermal methods. Mean-absolute-error (MAE), peak-signal-to-noise-ratio (PSNR), and structural similarity (SSIM) were used as metrics to evaluate the performance of the models. First, a super-resolution image (SR) was obtained using the traditional interpolation method used in computer vision. In the SR image at low magnification, the shape of the nanomaterial improved. However, the SR images at medium and high magnification failed to show the characteristics of the lattice of the nanomaterials. Second, to obtain a SR image, the deep learning model includes a residual network which reduces the loss of spatial information in the convolutional process of obtaining a feature map. In the process of optimizing the deep learning model, it was confirmed that the performance of the model improved as the number of data increased. In addition, by optimizing the deep learning model using the loss function, including MAE and SSIM at the same time, improved results of the nanomaterial lattice in SR images were achieved at medium and high magnifications. The final proposed deep learning model used four residual blocks to obtain the characteristic map of the low-resolution image, and the super-resolution image was completed using Upsampling2D and the residual block three times.

• 천문학회보
• /
• 제46권1호
• /
• pp.42.2-43
• /
• 2021

We address a question of what are three main channels that can best translate other channels in ultraviolet (UV) and extreme UV (EUV) observations. For this, we compare the image translations among the nine channels of the Atmospheric Imaging Assembly on the Solar Dynamics Observatory using a deep learning model based on conditional generative adversarial networks. In this study, we develop 170 deep learning models: 72 models for single-channel input, 56 models for double-channel input, and 42 models for triple-channel input. All models have a single-channel output. Then we evaluate the model results by pixel-to-pixel correlation coefficients (CCs) within the solar disk. Major results from this study are as follows. First, the model with 131 Å shows the best performance (average CC = 0.84) among single-channel models. Second, the model with 131 and 1600 Å shows the best translation (average CC = 0.95) among double-channel models. Third, among the triple-channel models with the highest average CC (0.97), the model with 131, 1600, and 304 Å is suggested in that the minimum CC (0.96) is the highest. Interestingly they are representative coronal, photospheric, and chromospheric lines, respectively. Our results may be used as a secondary perspective in addition to primary scientific purposes in selecting a few channels of an UV/EUV imaging instrument for future solar satellite missions.

• 한국멀티미디어학회논문지
• /
• 제25권11호
• /
• pp.1653-1671
• /
• 2022

Crop diseases affect crop production, more than 30 billion USD globally. We proposed a classification study of crop species and diseases using deep learning algorithms for corn, cucumber, pepper, and strawberry. Our study has three steps of species classification, disease detection, and disease classification, which is noteworthy for using captured images without additional processes. We designed deep learning approach of deep learning convolutional neural networks based on Mask R-CNN model to classify crop species. Inception and Resnet models were presented for disease detection and classification sequentially. For classification, we trained Mask R-CNN network and achieved loss value of 0.72 for crop species classification and segmentation. For disease detection, InceptionV3 and ResNet101-V2 models were trained for nodes of crop species on 1,500 images of normal and diseased labels, resulting in the accuracies of 0.984, 0.969, 0.956, and 0.962 for corn, cucumber, pepper, and strawberry by InceptionV3 model with higher accuracy and AUC. For disease classification, InceptionV3 and ResNet 101-V2 models were trained for nodes of crop species on 1,500 images of diseased label, resulting in the accuracies of 0.995 and 0.992 for corn and cucumber by ResNet101 with higher accuracy and AUC whereas 0.940 and 0.988 for pepper and strawberry by Inception.

• 융합보안논문지
• /
• 제22권4호
• /
• pp.129-133
• /
• 2022

딥러닝 프레임워크는 현재에도 계속해서 발전되어 가고 있으며, 다양한 프레임워크들이 존재한다. 딥러닝의 대표적인 프레임워크는 TensorFlow, PyTorch, Keras 등이 있다. 딥러님 프레임워크는 이미지 학습을 통해 이미지 분류에서의 최적화 모델을 이용한다. 본 논문에서는 딥러닝 이미지 인식 분야에서 가장 많이 사용하고 있는 TensorFlow와 PyTorch 프레임워크를 활용하여 이미지 학습을 진행하였으며, 이 과정에서 도출한 결과를 비교 분석하여 최적화된 프레임워크을 알 수 있었다.

• 한국멀티미디어학회논문지
• /
• 제21권2호
• /
• pp.98-107
• /
• 2018

In this paper, we propose a method to classify the skeletal maturity with a small amount of hand wrist X-ray image using deep learning-based meta-learning. General deep-learning techniques require large amounts of data, but in many cases, these data sets are not available for practical application. Lack of learning data is usually solved through transfer learning using pre-trained models with large data sets. However, transfer learning performance may be degraded due to over fitting for unknown new task with small data, which results in poor generalization capability. In addition, medical images require high cost resources such as a professional manpower and mcuh time to obtain labeled data. Therefore, in this paper, we use meta-learning that can classify using only a small amount of new data by pre-trained models trained with various learning tasks. First, we train the meta-model by using a separate data set composed of various learning tasks. The network learns to classify the bone maturity using the bone maturity data composed of the radiographs of the wrist. Then, we compare the results of the classification using the conventional learning algorithm with the results of the meta learning by the same number of learning data sets.

• Journal of Information Processing Systems
• /
• 제19권5호
• /
• pp.673-687
• /
• 2023

Convolutional neural network-based deep learning technology is the most commonly used in image identification, but it requires large-scale data for training. Therefore, application in specific fields in which data acquisition is limited, such as in the military, may be challenging. In particular, the identification of ground weapon systems is a very important mission, and high identification accuracy is required. Accordingly, various studies have been conducted to achieve high performance using small-scale data. Among them, the ensemble method, which achieves excellent performance through the prediction average of the pre-trained models, is the most representative method; however, it requires considerable time and effort to find the optimal combination of ensemble models. In addition, there is a performance limitation in the prediction results obtained by using an ensemble method. Furthermore, it is difficult to obtain the ensemble effect using models with imbalanced classification accuracies. In this paper, we propose a transfer learning-based feature fusion technique for heterogeneous models that extracts and fuses features of pre-trained heterogeneous models and finally, fine-tunes hyperparameters of the fully connected layer to improve the classification accuracy. The experimental results of this study indicate that it is possible to overcome the limitations of the existing ensemble methods by improving the classification accuracy through feature fusion between heterogeneous models based on transfer learning.

• International Journal of Internet, Broadcasting and Communication
• /
• 제15권2호
• /
• pp.133-143
• /
• 2023

In this study, we propose a big data set of strawberry pests collected directly for diagnosis model learning and an automatic pest diagnosis model architecture based on deep learning. First, a big data set related to strawberry pests, which did not exist anywhere before, was directly collected from the web. A total of more than 12,000 image data was directly collected and classified, and this data was used to train a deep learning model. Second, the deep-learning-based automatic pest diagnosis module is a module that classifies what kind of pest or disease corresponds to when a user inputs a desired picture. In particular, we propose a model architecture that can optimally classify pests based on a convolutional neural network among deep learning models. Through this, farmers can easily identify diseases and pests without professional knowledge, and can respond quickly accordingly.

• International Journal of Computer Science & Network Security
• /
• 제21권9호
• /
• pp.223-229
• /
• 2021

Classification of different blood cell types is an essential task for human's medical treatment. The white blood cells have different types of cells. Counting total White Blood Cells (WBC) and differential of the WBC types are required by the physicians to diagnose the disease correctly. This paper used transfer learning methods to the pre-trained deep learning models to classify different WBCs. The best pre-trained model was Inception ResNetV2 with Adam optimizer that produced classification accuracy of 98.4% for the dataset comprising four types of WBCs.

• 전자통신동향분석
• /
• 제34권2호
• /
• pp.40-50
• /
• 2019

Considerable accuracy improvements in deep learning have recently been achieved in many applications that require large amounts of computation and expensive memory. However, recent advanced techniques for compacting and accelerating the deep learning model have been developed for deployment in lightweight devices with constrained resources. Lightweight deep learning techniques can be categorized into two schemes: lightweight deep learning algorithms (model simplification and efficient convolutional filters) in nature and transferring models into compact/small ones (model compression and knowledge distillation). In this report, we briefly summarize various lightweight deep learning techniques and possible research directions.