• 한국항공운항학회지
• /
• 제23권1호
• /
• pp.120-125
• /
• 2015

Passengers and Airlines wish neither delay nor cancellation due to aircraft defects. However, about 1 delay or cancellation case occurs out of 100 departures worldwide whereas 1 quarter case does in Korean domestic industry. Independent LCC carriers in Korea have almost double case. Most cases are recovered by replacing aircraft components. Airlines have prepared the spare components based on the reliability data by manufacturers to rectify defects or perform preventive maintenances. The total value for initial spares including engine is 40% of the aircraft price when they operate less than 5 aircraft. The more airlines operate the aircraft, the less the ratio of the investment for spares reflecting the economy of scale. This study intends to suggest how to improve the efficiencies as well as the safety of LCC throughout parts pooling including engines.

• 한국가축번식학회지
• /
• 제8권1호
• /
• pp.46-55
• /
• 1984

The purpose of the experiment was to clarify morphologically normal growth pattern of the ductus deference in accordance with the sex maturity of meat-type cockerels. 1. Diameter of lumens in u, pp.r, mid and lower parts of ductus deferens, the most conspicuous enlargement of lumen was observed in the lower part. Heights of epithelial layers of ductus deferens showed abrupt growth at 12 weeks of age with subsequent gradual growth in all the part of u, pp.r, mid and lower, and heights of those at 30 weeks were a, pp.oximately 4 times as large in the u, pp.r and mid parts and 5 times as large in the lower part in contrast to those at 4 weeks of age. Thickness of muscular layer of ductus showed gradual growth in contrast with the diameter of lumen and height of epithelial layer, showing 1.3 times as large in the u, pp.r part, 1.6 times in the mid part and 1.9 times in the lower part at 30 weeks of age in contrast to the thickness at 4 weeks of age. 2. Within 10 weeks after hatching, lining cells of ductus deferens were mainly composed of round cells and columnar cells in simple columnar epithelium. During 10th to 20th week, the lining cells were mainly composed of high columnar cells and round cells in pseudostratified epithelium. From 22nd week, the lining cells were composed of pseudostratified columnar cells. Whereas round cells disa, pp.ared gradually. Enlargement of lumen and pooling of sperms in ductus deferens coincided with the maturation of seminiferous tubules. 3. In simple correlation between the values of testis weight and the values from various measurements in the ductus deferens, there was significant correlation coefficient with each other. 4. In the India ink absorption test, India ink granules were not absorbed on the epithelium of the ductus deferens, but the granules reactive to acid phosphatase a, pp.ared in a line on the free border of each parts of the ductus deferens. The granules reactive to alkaline phosphatase were noted on the luminal border of ductus deferens mainly, but weak reaction showed than acid phophatase were a, pp.ared. The granules reactive to PAS were a, pp.ared mostly near on the free border of hte epithelial cells of ductus deferens. 5. Number of sperm, Indes of sperm vitality and MRT in the different parts of ductus deferens were tended to be somewhat dominant in the mid and lower parts than in u, pp.r part, even though not significant in the statistical analysis. Ratio of sperm abnormality was tended to be relatively high in the u, pp.r part too, and in the sperm of abnormality blunted head was less in number significantly in the mid and lower part than in the u, pp.r part.

• 한국멀티미디어학회논문지
• /
• 제21권12호
• /
• pp.1425-1430
• /
• 2018

In this paper, we propose a CNN based deep learning algorithm for semantic segmentation of images. In order to improve the accuracy of semantic segmentation, we combined pixel level object classification and image level object classification. The image level object classification is used to accurately detect the characteristics of an image, and the pixel level object classification is used to indicate which object area is included in each pixel. The proposed network structure consists of three parts in total. A part for extracting the features of the image, a part for outputting the final result in the resolution size of the original image, and a part for performing the image level object classification. Loss functions exist for image level and pixel level classification, respectively. Image-level object classification uses KL-Divergence and pixel level object classification uses cross-entropy. In addition, it combines the layer of the resolution of the network extracting the features and the network of the resolution to secure the position information of the lost feature and the information of the boundary of the object due to the pooling operation.

• Journal of Information Processing Systems
• /
• 제19권6호
• /
• pp.842-857
• /
• 2023

Currently, most sentiment classification models on microblogging platforms analyze sentence parts of speech and emoticons without comprehending users' emotional inclinations and grasping moral nuances. This study proposes a hybrid sentiment analysis model. Given the distinct nature of microblog comments, the model employs a combined stop-word list and word2vec for word vectorization. To mitigate local information loss, the TextCNN model, devoid of pooling layers, is employed for local feature extraction, while BiLSTM is utilized for contextual feature extraction in deep learning. Subsequently, microblog comment sentiments are categorized using a classification layer. Given the binary classification task at the output layer and the numerous hidden layers within BiLSTM, the Tanh activation function is adopted in this model. Experimental findings demonstrate that the enhanced TextCNN-BiLSTM model attains a precision of 94.75%. This represents a 1.21%, 1.25%, and 1.25% enhancement in precision, recall, and F1 values, respectively, in comparison to the individual deep learning models TextCNN. Furthermore, it outperforms BiLSTM by 0.78%, 0.9%, and 0.9% in precision, recall, and F1 values.

• 한국컴퓨터정보학회논문지
• /
• 제29권9호
• /
• pp.1-7
• /
• 2024

본 논문은 MNIST 데이터셋을 활용한 손글씨 숫자 인식에서 합성곱 신경망(CNN)과 배치정규화(BN)를 결합한 모델을 제안한다. LeCun et al.의 LeNet-5 모델의 성과를 뛰어넘는 것을 목표로 6계층 신경망 구조를 설계하였다. 제안된 모델은 28×28 픽셀 이미지를 입력으로 받아 합성곱, 맥스 풀링, 완전연결계층을 거쳐 처리하며, 특히 배치정규화계층을 도입하여 학습 안정성과 성능을 향상시켰다. 실험에서는 60,000개의 훈련 이미지와 10,000개의 테스트 이미지를 사용하였으며, Momentum 최적화 알고리즘을 적용하였다. 모델 구성에서는 30개의 필터, 필터 사이즈 5×5, 패딩 0, 스트라이드 1을 사용하였고, ReLU 활성화 함수를 채택하였다. 훈련 과정에서는 미니배치 사이즈 100, 총 20 에포크, 학습률 0.1로 설정하였다. 결과적으로 제안된 모델은 99.22%의 테스트 정확도를 달성하여 LeNet-5의 99.05%를 상회하였으며, F1-score 0.9919를 기록하여 모델의 성능을 입증하였다. 또한, 본 논문에서 제안한 6계층 모델은 LeCun et al.의 LeNet-5(7계층 모델)와 Ji, Chun and Kim(10계층 모델)이 제안한 모델보다 더 단순한 구조로 모델의 효율성을 강조하였다. 본 연구의 결과는 AI 비전 검사기 등 실제 산업 응용에서 활용 가능성을 보여주며, 특히 스마트팩토리에서 부품의 불량 상태를 판별하는 데 효과적으로 적용될 수 있을 것으로 기대된다.

• 농업과학연구
• /
• 제47권4호
• /
• pp.1109-1122
• /
• 2020

The aim of this study was to develop a marbling classification and prediction model using small parts of sirloin images based on a deep learning algorithm, namely, a convolutional neural network (CNN). Samples were purchased from a commercial slaughterhouse in Korea, images for each grade were acquired, and the total images (n = 500) were assigned according to their grade number: 1++, 1+, 1, and both 2 & 3. The image acquisition system consists of a DSLR camera with a polarization filter to remove diffusive reflectance and two light sources (55 W). To correct the distorted original images, a radial correction algorithm was implemented. Color images of sirloins of Hanwoo (mixed with feeder cattle, steer, and calf) were divided and sub-images with image sizes of 161 × 161 were made to train the marbling prediction model. In this study, the convolutional neural network (CNN) has four convolution layers and yields prediction results in accordance with marbling grades (1++, 1+, 1, and 2&3). Every single layer uses a rectified linear unit (ReLU) function as an activation function and max-pooling is used for extracting the edge between fat and muscle and reducing the variance of the data. Prediction accuracy was measured using an accuracy and kappa coefficient from a confusion matrix. We summed the prediction of sub-images and determined the total average prediction accuracy. Training accuracy was 100% and the test accuracy was 86%, indicating comparably good performance using the CNN. This study provides classification potential for predicting the marbling grade using color images and a convolutional neural network algorithm.

• 대한수의학회지
• /
• 제23권2호
• /
• pp.137-148
• /
• 1983

This study was made for the better information of the male reproductive system on the meat-type drake, Cherry Belly X White Golden. The epithelium of ductules of epididymal region and deferent duct were observed histologically and histochemically with the progress of their development. India-ink absorbability on the luminal epithelium was also investigated after the administration of India-ink. The results are as follows; 1. Rete testis and various round ductules in immature form appeared in epididymis within 6 weeks after hatching, and simple cuboidal and simple columnar epithelium were found in the epithelia of the ductules within 8 weeks after hatching. Larger ductules were found on epididymal surface which was in the developing stage near to the immature efferent ductule. From 10th to 20th week, various ductules appeared in epididymis, and developing form of efferent ductules were much more increased on epididymal surface. The luminal epithelium of the ductules were composed of ciliated simple columnar and pseudostratified ciliated columnar cells. At the same time, deferent duct appeared. From the 21th week, various ductules in epididymis became abruptly matured. Lumen of rete testis was lined by simple squamous or simple cuboidal epithelium, and that of efferent ductules, having many folds and being larger than any others were lined by pseudostratified ciliated columnar epithelium in which ciliated columnar cells, non-ciliated cells(clear cells) and basal cells were noted. Connecting tubules of star shaped lumen were composed of pseudostratified ciliated columnar epithelium in which ciliated columnar cells, nonciliated cells, and basal cells were observed. The luminal surface of epididymal ducts was smooth and has thick pseudostratified columnar epithelium which was composed of high columnar cells and basal cells. From 26th week after hatching, sperm pooling was started in various ductules. 2. From 4th to 10th week, simple cuboidal epithelium of deferent duct transformed to simple columnar epithelium with the progress of aging. At the basement of epithelium, clear round cells were noted. From 12th to 20th week, high columnar cells with enlongated nucleus were noted on the luminal border of deferent ducts, forming folds of pseuclostratified columnar epithelium. From 20th week, the deferent duct started to have septa in it's lumen and composed mainly of pseudostratified columnar epithelium, and round cells disappeared. From 20th week, the lumen diameter of deferent duct became wider with the progress of aging, but there was no difference among the values of lumen diameter in upper, middle, and lower part of deferent ducts. At 26th week, the pooling period of sperms in deferent ducts, the lumen diameter became rapidly widen, especially in the lower part of deferent ducts. Thickness of muscular layer of ductus deferens showed gradual growth within 24 weeks but did abrupt thickening from 26th week. 3. Saliva resistant PAS granules were dotted on the top of nucleus in efferent ductules epithelium but the amount of the granules were little in the connecting ductules's epithelium. The granules reactive to acid phosphatase were abundant in the some epithelial cells of efferent ductules and connecting ductules, especially above the nucleus of cells. The granules reactive to alkaline phosphatase were noted on the luminal border of efferent ductules. Parts of free border of efferent ductules and middle portion of deferent ducts were stained slightly by alcian blue technique. India ink granules were found mainly in the epithelium of efferent ductules but were few in that of connecting ductules.