• Annual Conference on Human and Language Technology
• /
• 2021.10a
• /
• pp.391-395
• /
• 2021

텍스트 스타일 변환은 입력 스타일(source style)로 쓰여진 텍스트의 내용(content)을 유지하며 목적 스타일(target style)의 텍스트로 변환하는 문제이다. 텍스트 스타일 변환을 시퀀스 간 변환 문제(sequence-to-sequence)로 보고 기존 기계학습 모델을 이용해 해결할 수 있지만, 모델 학습에 필요한 각 스타일에 대응되는 병렬 말뭉치를 구하기 어려운 문제점이 있다. 따라서 최근에는 비병렬 말뭉치를 이용해 텍스트 스타일 변환을 수행하는 방법들이 연구되고 있다. 이 연구들은 주로 인코더-디코더 구조의 생성 모델을 사용하기 때문에 입력 문장이 가지고 있는 내용이 누락되거나 다른 내용의 문장이 생성될 수 있는 문제점이 있다. 본 논문에서는 마스크 언어 모델(masked language model)을 이용해 입력 텍스트의 내용을 유지하면서 원하는 스타일로 변경할 수 있는 텍스트 스타일 변환 방법을 제안하고 한국어 긍정-부정, 채팅체-문어체 변환에 적용한다.

• Journal of the Korean Society of Costume
• /
• v.18
• /
• pp.321-337
• /
• 1992

Fashions are constantly subject to change, so that forecasting trends in fashion is very difficult and the forecasting is very important. The reason is that forecasting is identifying a hot trend for the next days' business and projecting the future. The purpose of this study is to suggest the realistic forecasting theory identify fashion trends at present, further forecast fashion trends to 1995. For this study, historical models (th historical resurrection model, the historical continuity model) and the art movement model among the behavioral science models which was suggested in 1986 by Geoge B.Sproles were selected. According to this study, was exhibited. This curve was resulted from relation between 1980s fashion trends and the behavioral science models of fashion. Fashion trends to 1995 through will be followings; 1. Body Conscious look as one of the characteristics of modern fashion will bi continued throughout 1990s' fashion trends. 2. The attraction of Mini style was started from the middle of 1980s, recently gets near micro mini and super shouts, These are representative of the historical continuity model. 3. Mini look as revivals of '60s in early 1990s fashion trends. This phenomenon identifies that "Fashion is repetition" 4. When Mini look and super shorts reach, will not progress any more. In this time, will bi seen transfer of style be using classic style. Tat is modern classic will be appeared. also this condition is same from late 1960s to early 1970s. In view of the historical base, the action of the varieted pants will be prominent. 5. Fashion trends which passed through this process will transfer from young look to elegant look. In this time, designers will retrospect '30s and '50s fashion trends, and then will design modern feminine look. 6. Ethnic look, the fashion trend after that condition will be accelated feminine look. This look with the feminine look will reveal rich, colorful, elegance feminine image. 7. In 1980s, activity and function in clothes were demanded. The clothes which are considered these condition will be demanded continuously. 8. The variety of fashion will appear as mixmode and new coordinate look in a clothes simultaneously in fashion of '90s. And these looks will creat mixed beauty. 9. Pop art, op art, graphic art will be not only development of textile but also movement of fashion soul in 1990s. The fashion behavior will be continuously scientificated.

• KIPS Transactions on Software and Data Engineering
• /
• v.10 no.12
• /
• pp.579-586
• /
• 2021

Time series forecasting refers to predicting future time information based on past time information. Accurately predicting future information is crucial because it is used for establishing strategies or making policy decisions in various fields. Recently, a transformer model has been mainly studied for a time series prediction model. However, the existing transformer model has a limitation in that it has an auto-regressive structure in which the output result is input again when the prediction sequence is output. This limitation causes a problem in that accuracy is lowered when predicting a distant time point. This paper proposes a sequential decoding model focusing on the style transformation technique to handle these problems and make more precise time series forecasting. The proposed model has a structure in which the contents of past data are extracted from the transformer-encoder and reflected in the style-based decoder to generate the predictive sequence. Unlike the decoder structure of the conventional auto-regressive transformer, this structure has the advantage of being able to more accurately predict information from a distant view because the prediction sequence is output all at once. As a result of conducting a prediction experiment with various time series datasets with different data characteristics, it was shown that the model presented in this paper has better prediction accuracy than other existing time series prediction models.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.17 no.2
• /
• pp.291-298
• /
• 2022

A generative adversarial network (GAN) is a network in which two internal neural networks (generative network and discriminant network) learn while competing with each other. The generator creates an image close to reality, and the delimiter is programmed to better discriminate the image of the constructor. This technology is being used in various ways to create, transform, and restore the entire image X into another image Y. This paper describes a method that can be forged into another object naturally, after extracting only a partial image from the original image. First, a new image is created through the previously trained DCGAN model, after extracting only a partial image from the original image. The original image goes through a process of naturally combining with, after re-styling it to match the texture and size of the original image using the overall style transfer technique. Through this study, the user can naturally add/transform the desired object image to a specific part of the original image, so it can be used as another field of application for creating fake images.

• Annual Conference on Human and Language Technology
• /
• 2018.10a
• /
• pp.567-569
• /
• 2018

한국어의 경어체는 종결어미에 따라 구분하고, 서로 다른 경어체는 각각 고유한 경어 강도가 있다. 경어체 간의 어체 변환은 규칙기반으로 진행되어 왔다. 본 논문은 어체 변환을 위한 규칙 정의의 번거로움을 줄이고 어체 변환 데이터만을 사용한 심층 학습 기반의 어체 변환 방법을 제안한다. 본 연구는 '해요체-합쇼체' 쌍의 병렬 데이터를 이용하여 Attention-based Sequence-to-Sequence 모델을 바탕으로 한 어체 변환 모델을 학습하였다. 해당 모델을 학습하고 실험하였을 때, 정확도 91%의 우수한 성과를 얻을 수 있었다.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.15 no.6
• /
• pp.2115-2127
• /
• 2021

Although significant progress has been made in synthesizing visually realistic face images by Generative Adversarial Networks (GANs), there still lacks effective approaches to provide fine-grained control over the generation process for semantic facial attribute editing. In this work, we propose a novel cross channel self-attention based generative adversarial network (CCA-GAN), which weights the importance of multiple channels of features and archives pixel-level feature alignment and conversion, to reduce the impact on irrelevant attributes while editing the target attributes. Evaluation results show that CCA-GAN outperforms state-of-the-art models on the CelebA dataset, reducing Fréchet Inception Distance (FID) and Kernel Inception Distance (KID) by 15~28% and 25~100%, respectively. Furthermore, visualization of generated samples confirms the effect of disentanglement of the proposed model.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2021.11a
• /
• pp.741-744
• /
• 2021

본 논문은 사전 학습된 심층생성모델을 기반으로 가수 별 가사의 특성을 반영하여 새로운 가사를 생성하는 모델을 소개한다. 베이스 모델로 한국어 사전 학습 모델 KoGPT-2 를 사용하였으며, 총 가수 10 명의 노래 823 곡을 수집하여 미세조정 기법을 바탕으로 학습하였다. 특히, 가수 별로 구분한 가사를 학습 데이터로 구축하여, 가수 별로 독특하게 나타나는 가사 스타일이 전이되도록 하였다. 가수의 이름과 시작 단어를 입력으로 주고 작사를 수행한 실험 결과, (i) 가수 별로 생성되는 가사의 어휘와 스타일이 그 가수의 기존 곡들의 가사와 유사함을 확인하였고, (ii) 작사 결과 가수 별 차이를 확인하였다. 추후 설문을 통해, 개별 가수들의 가사와 생성된 가사의 어휘와 스타일 유사성을 확인하고, 가수 별 차이 또한 확인하고자 한다.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.18 no.2
• /
• pp.311-326
• /
• 2024

The rapid development of neural network technology promotes the neural network model driven by big data to overcome the texture effect of complex objects. Due to the limitations in complex scenes, it is necessary to establish custom template matching and apply it to the research of many fields of computational vision technology. The dependence on high-quality small label sample database data is not very strong, and the machine learning system of deep feature connection to complete the task of texture effect inference and speculation is relatively poor. The style transfer algorithm based on neural network collects and preserves the data of patterns, extracts and modernizes their features. Through the algorithm model, it is easier to present the texture color of patterns and display them digitally. In this paper, according to the texture effect reasoning of custom template matching, the 3D visualization of the target is transformed into a 3D model. The high similarity between the scene to be inferred and the user-defined template is calculated by the user-defined template of the multi-dimensional external feature label. The convolutional neural network is adopted to optimize the external area of the object to improve the sampling quality and computational performance of the sample pyramid structure. The results indicate that the proposed algorithm can accurately capture the significant target, achieve more ablation noise, and improve the visualization results. The proposed deep convolutional neural network optimization algorithm has good rapidity, data accuracy and robustness. The proposed algorithm can adapt to the calculation of more task scenes, display the redundant vision-related information of image conversion, enhance the powerful computing power, and further improve the computational efficiency and accuracy of convolutional networks, which has a high research significance for the study of image information conversion.

• Journal of the Korea Society of Computer and Information
• /
• v.29 no.3
• /
• pp.11-20
• /
• 2024

Among the generative models in Artificial Intelligence (AI), especially Generative Adversarial Network (GAN) has been successful in various applications such as image processing, density estimation, and style transfer. While the GAN models including Conditional GAN (CGAN), CycleGAN, BigGAN, have been extended and improved, researchers face challenges in real-world applications in specific domains such as disaster simulation, healthcare, and urban planning due to data scarcity and unstable learning causing Image distortion. This paper proposes a new progressive learning methodology called Progressive Step Training (PST) based on the Auxiliary Classifier GAN (ACGAN) that discriminates class labels, leveraging the progressive learning approach of the Progressive Growing of GAN (PGGAN). The PST model achieves 70.82% faster stabilization, 51.3% lower standard deviation, stable convergence of loss values in the later high resolution stages, and a 94.6% faster loss reduction compared to conventional methods.

• Journal of the Korean Solar Energy Society
• /
• v.33 no.4
• /
• pp.70-79
• /
• 2013

These day cities experience serious climatic changes due to environmental load caused by disturbance in the circulation systems of water resources and energy. As technological improvement to respond to various climatic changes and disasters are also requested in the field of construction, inter-disciplinary studies linked to the establishment of sustainable environmental control and energy systems is required in a consilient perspective. This study aims to infer correlations in the impact of environmental changes caused by rooftop greening system on the photovoltaic power generation efficiency through computer simulation in an integrated perspective. By doing so, it seeks to provide basic study for developing a photovoltaic system integrated with building revegetation that is sustainable in environmental and resource aspects. A simulation showed that, in the case of sunshine hours in June, the green surface indicated temperature lowering effects of $9.19^{\circ}C$ on average compared to the non-green surface and temperature was $9.81^{\circ}C$ lower. Due to such greening effects, at the highest sunlight timepoint in June, Pmpp improved 119W and heat loss rate dropped 7.8%.