• Journal of the Korea Convergence Society
• /
• v.12 no.3
• /
• pp.1-7
• /
• 2021

Since Neural Machine Translation (NMT) uses only a limited number of words, there is a possibility that words that are not registered in the dictionary will be entered as input. The proposed method to alleviate this Out of Vocabulary (OOV) problem is Subword Tokenization, which is a methodology for constructing words by dividing sentences into subword units smaller than words. In this paper, we deal with general subword tokenization algorithms. Furthermore, in order to create a vocabulary that can handle the infinite conjugation of Korean adjectives and verbs, we propose a new methodology for subword tokenization training by separating the Jongsung(coda) from Korean syllables (consisting of Chosung-onset, Jungsung-neucleus and Jongsung-coda). As a result of the experiment, the methodology proposed in this paper outperforms the existing subword tokenization methodology.

• Smart Media Journal
• /
• v.10 no.1
• /
• pp.63-69
• /
• 2021

When a person sees a sentence and understands the sentence, the person understands the sentence by reminiscent of the main word in the sentence as an image. Text-to-image is what allows computers to do this associative process. The previous deep learning-based text-to-image model extracts text features using Convolutional Neural Network (CNN)-Long Short Term Memory (LSTM) and bi-directional LSTM, and generates an image by inputting it to the GAN. The previous text-to-image model uses basic embedding in text feature extraction, and it takes a long time to train because images are generated using several modules. Therefore, in this research, we propose a method of extracting features by using the attention mechanism, which has improved performance in the natural language processing field, for sentence embedding, and generating an image by inputting the extracted features into the GAN. As a result of the experiment, the inception score was higher than that of the model used in the previous study, and when judged with the naked eye, an image that expresses the features well in the input sentence was created. In addition, even when a long sentence is input, an image that expresses the sentence well was created.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.25 no.3
• /
• pp.389-395
• /
• 2021

In this paper, we propose a target image exchange model to improve performance of the object tracking algorithm based on a Siamese network. The object tracking algorithm based on the Siamese network tracks the object by finding the most similar part in the search image using only the target image specified in the first frame of the sequence. Since only the object of the first frame and the search image compare similarity, if tracking fails once, errors accumulate and drift in a part other than the tracked object occurs. Therefore, by designing a CNN(Convolutional Neural Network) based model, we check whether the tracking is progressing well, and the target image exchange timing is defined by using the score output from the Siamese network-based object tracking algorithm. The proposed model is evaluated the performance using the VOT-2018 dataset, and finally achieved an accuracy of 0.611 and a robustness of 22.816.

• Korean Journal of Remote Sensing
• /
• v.36 no.6_2
• /
• pp.1485-1492
• /
• 2020

It has been more than 20 years since the launch of KOMPSAT-1, and so far, a total of 5 satellites have been successfully launched. Until now, KOMPSAT has been used in various fields, including the production of various thematic maps, land change, environmental analysis, and marine monitoring. Many researchers have conducted research to process, analyze, and utilize KOMPSAT images. According to the national space development plan, the KOMPSAT series will be continuously developed to meet the demand for satellite images at the national level. If the ultimate purpose of satellite development is to utilize acquired images, systematic research to effectively utilize the developed satellites should be followed. This special issue introduces the recently conducted research on the use of KOMPSAT images.

• Journal of the Korea Society for Simulation
• /
• v.29 no.4
• /
• pp.1-7
• /
• 2020

Early detection and classification of crop diseases play significant role to help farmers to reduce disease spread and to increase agricultural productivity. Recently, many researchers have used deep learning techniques like convolutional neural network (CNN) classifier for crop disease inspection with dataset of crop leaf images (e.g., PlantVillage dataset). These researches present over 90% of classification accuracy for crop diseases, but they have ability to detect only the pre-trained diseases. This paper proposes an efficient disease inspection CNN model for new crops not used in the pre-trained model. First, we present a benchmark crop disease classifier (CDC) for the crops in PlantVillage dataset using VGG16. Then we build a modified crop disease classifier (mCDC) to inspect diseases for untrained crops. The performance evaluation results show that the proposed model outperforms the benchmark classifier.

• Smart Media Journal
• /
• v.10 no.1
• /
• pp.39-46
• /
• 2021

This study is about a method of extracting a summary from a news article in consideration of the importance of each sentence constituting the article. We propose a method of calculating sentence importance by extracting the probabilities of topic sentence, similarity with article title and other sentences, and sentence position as characteristics that affect sentence importance. At this time, a hypothesis is established that the Topic Sentence will have a characteristic distinct from the general sentence, and a deep learning-based classification model is trained to obtain a topic sentence probability value for the input sentence. Also, using the pre-learned ELMo language model, the similarity between sentences is calculated based on the sentence vector value reflecting the context information and extracted as sentence characteristics. The topic sentence classification performance of the LSTM and BERT models was 93% accurate, 96.22% recall, and 89.5% precision, resulting in high analysis results. As a result of calculating the importance of each sentence by combining the extracted sentence characteristics, it was confirmed that the performance of extracting the topic sentence was improved by about 10% compared to the existing TextRank algorithm.

• Journal of The Korean Association of Information Education
• /
• v.25 no.3
• /
• pp.471-481
• /
• 2021

The purpose of this study is to analyze the effectiveness of AI convergence education program. To this end, the "Elementary Science AI Convergence Education Program for Machine Learning" developed in previous research were taught to elementary school students in the fourth to sixth grades in eight times. The quantitative changes of each factor were analyzed by R program, and the effectiveness of education was analyzed by Pearson correlation and paired samples t-test. As a result, there is a deep correlation between "Attitude to AI technology, Scientific preference and STEAM Literacy" and technical average has improved in many factors. Therefore, AI convergence education program is meaningful in terms of education, and if AI education and AI convergence education are implemented into the primary formal education curriculum, they will have a positive effect.

• Journal of the Korea Society of Computer and Information
• /
• v.26 no.8
• /
• pp.31-37
• /
• 2021

In this paper, we propose a method to effectively search by automatically classifying scenes in which specific images such as pitching or swing appear in baseball game images using a CNN(Convolution Neural Network) model. In addition, we propose a video scene search system that links the classification results of specific motions and game records. In order to test the efficiency of the proposed system, an experiment was conducted to classify the Korean professional baseball game videos from 2018 to 2019 by specific scenes. In an experiment to classify pitching scenes in baseball game images, the accuracy was about 90% for each game. And in the video scene search experiment linking the game record by extracting the scoreboard included in the game video, the accuracy was about 80% for each game. It is expected that the results of this study can be used effectively to establish strategies for improving performance by systematically analyzing past game images in Korean professional baseball games.

• Phonetics and Speech Sciences
• /
• v.13 no.4
• /
• pp.47-53
• /
• 2021

Autoregressive Text-to-Speech (TTS) models suffer from inference instability and slow inference speed. Inference instability occurs when a poorly predicted sample at time step t affects all the subsequent predictions. Slow inference speed arises from a model structure that forces the predicted samples from time steps 1 to t-1 to predict the sample at time step t. In this study, an end-to-end non-autoregressive fast text-to-speech model is suggested as a solution to these problems. The results of this study show that this model's Mean Opinion Score (MOS) is close to that of Tacotron 2 - WaveNet, while this model's inference speed and stability are higher than those of Tacotron 2 - WaveNet. Further, this study aims to offer insight into the improvement of non-autoregressive models.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.26 no.1
• /
• pp.24-29
• /
• 2022

Semantic role labeling task used in various natural language processing fields, such as information extraction and question answering systems, is the task of identifying the arugments for a given sentence and predicate. Predicate used as semantic role labeling input are extracted using lexical analysis results such as POS-tagging, but the problem is that predicate can't extract all linguistic patterns because predicate in korean language has various patterns, depending on the meaning of sentence. In this paper, we propose a korean predicate recognition method using neural network model with pre-trained embedding models and lexical features. The experiments compare the performance on the hyper parameters of models and with or without the use of embedding models and lexical features. As a result, we confirm that the performance of the proposed neural network model was 92.63%.