• Journal of Information Processing Systems
• /
• v.13 no.4
• /
• pp.928-940
• /
• 2017

State-of-the-art speaker recognition systems may work better for the English language. However, if the same system is used for recognizing those who speak different languages, the systems may yield a poor performance. In this work, the decisions of a Gaussian mixture model-universal background model (GMM-UBM) and a learning vector quantization (LVQ) are combined to improve the recognition performance of a multilingual speaker identification system. The difference between these classifiers is in their modeling techniques. The former one is based on probabilistic approach and the latter one is based on the fine-tuning of neurons. Since the approaches are different, each modeling technique identifies different sets of speakers for the same database set. Therefore, the decisions of the classifiers may be used to improve the performance. In this study, multitaper mel-frequency cepstral coefficients (MFCCs) are used as the features and the monolingual and cross-lingual speaker identification studies are conducted using NIST-2003 and our own database. The experimental results show that the combined system improves the performance by nearly 10% compared with that of the individual classifier.

• Annual Conference of KIPS
• /
• 2019.05a
• /
• pp.455-457
• /
• 2019

In this paper we study the effect of decomposed spatiotemporal convolutions for action recognition in videos. Our motivation emerges from the empirical observation that spatial convolution applied on solo frames of the video provide good performance in action recognition. In this research we empirically show the accuracy of factorized convolution on individual frames of video for action classification. We take 3D ResNet-18 as base line model for our experiment, factorize its 3D convolution to 2D (Spatial) and 1D (Temporal) convolution. We train the model from scratch using Kinetics video dataset. We then fine-tune the model on UCF-101 dataset and evaluate the performance. Our results show good accuracy similar to that of the state of the art algorithms on Kinetics and UCF-101 datasets.

• 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.

• International Journal of Advanced Culture Technology
• /
• v.10 no.4
• /
• pp.210-218
• /
• 2022

Chinese Opera-themed woodcut New Year pictures in Yangjiabu are representative of Chinese New Year pictures. This research takes cultural research as a breakthrough point, analyzes and combs its Context, Representation, Coding & Decoding, and finds that: After the imitation and restatement of traditional Chinese opera, the opera-themed woodcut New Year pictures have formed a dynamic intertextual relationship with Chinese culture and traditional Chinese opera. The understanding of this intertextual relationship is not only an important way to study Chinese folk art but also one of the ideas to show the aesthetic spirit of it.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2023.05a
• /
• pp.165-165
• /
• 2023

Streamflow prediction is a critical task in water resources management and essential for planning and decision-making purposes. However, the streamflow prediction is challenging due to the complexity and non-linear nature of hydrological processes. The transfer learning is a powerful technique that enables a model to transfer knowledge from a source domain to a target domain, improving model performance with limited data in the target domain. In this study, we apply the transfer learning using the Informer model, which is a state-of-the-art deep learning model for streamflow prediction. The model was trained on a large-scale hydrological dataset in the source basin and then fine-tuned using a smaller dataset available in the target basin to predict the streamflow in the target basin. The results demonstrate that transfer learning using the Informer model significantly outperforms the traditional machine learning models and even other deep learning models for streamflow prediction, especially when the target domain has limited data. Moreover, the results indicate the effectiveness of streamflow prediction when knowledge transfer is used to improve the generalizability of hydrologic models in data-sparse regions.

• International Journal of Advanced Culture Technology
• /
• v.11 no.1
• /
• pp.192-197
• /
• 2023

David Salle, born in Norman, Oklahoma in 1952, is the most important and controversial artist of the new-expressionist painting in the United States. His paintings are featured by unique artistic language, and extremely diverse and very ground-breaking expression, which reflect the American society at that time. Besides, he often uses complex image combination with striking colors and forms to attract and touch viewers. In this paper, based on the American new-expressionism, the painting language of David Salle's works was analysed, in which I found that it can be more intuitively and more profoundly feel the development of new-expressionist painting context and David Salle's particularity, and more clearly realize under what the characteristics of the era David Salle created the unique artistic language. The study on David Salle's unique painting language is of important research value and reference significance for our current artistic creation.

• International Journal of Advanced Culture Technology
• /
• v.11 no.1
• /
• pp.204-211
• /
• 2023

Chinese figure painting is the earliest type of painting to be mature in the history of Chinese painting. After its peak in the Jin and Tang dynasties, Chinese figure painting turned into a low tide in the Ming and Qing dynasties. Under the influence of the May Fourth New Culture Movement's concept of "introducing the Western culture into China", realistic ink wash figure painting emerged, which was different from traditional painting. Realistic ink wash figure painting is a new kind of painting that combines traditional Chinese painting methods with concepts of Western scientific painting. With the mutual reconciliation of realistic modeling and traditional Chinese painting, it makes up for the shortcomings of traditional Chinese figure painting in reproducing life, thus forming the diversified development of Chinese figure painting. Through an in-depth interpretation of the fusion of Chinese realistic ink figure painting with Chinese and Western, this paper accurately grasps the characteristics of the painting language, which is an important guide to the creation of Chinese realistic ink painting today.

• International Journal of Advanced Culture Technology
• /
• v.11 no.1
• /
• pp.389-394
• /
• 2023

Based on intelligent visual information and 5G, this paper studies the intelligent visual communication of landscape bamboo buildings, and provides a new method of intelligent perception and interactive computing for the real world, which can represent, model, Perception and cognition; through the integration of virtual and real, the situational understanding of the human-machine-material fusion environment and the interaction with nature. The 5G network can well meet the combination of high-bandwidth uplink transmission and low-latency downlink control. At the same time, 5G-based AR intelligent inspection, remote operation and maintenance guidance, and machine vision inspection. Taking the bamboo building as an example, through field inspections to analyze tourism Bamboo buildings before and after development, and the intelligentization of bamboo buildings based on 5G and visual modeling.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2020.11a
• /
• pp.33-36
• /
• 2020

Virtual try on is getting interested from researchers these days because its application in online shopping. But single pose virtual try on is not enough, customer may want to see themselves in different pose. Multiple pose virtual try on is getting input as customer image, an in-shop cloth and a target pose, it will try to generate realistic customer wearing the in-shop cloth with the target pose. We first generate the target segmentation layout using conditional generative network (cGAN), and then the in-shop cloth are warped to fit the customer body in target pose. Finally, all the result will be combine using a Resnet-like network. We experiment and show that our method outperforms stage of the art.

• SUVANNABHUMI
• /
• v.1 no.2
• /
• pp.73-91
• /
• 2009

베트남 관음상, 특히 천 개의 팔과 눈이 있는 관음상은 일반적으로는 아시아에서, 구체적으로는 동남아에서 종교 혼합주의의 가장 명백한 증거로 여겨질 수 있다. 그것은 인도와 더불어 특히 중국모델에서 전래된 몇 가지 구체적인 증거를 베트남 관음상에서 발견할 수 있다. 많은 학자들은 시바를 관음상의 천개의 팔과 천개의 눈을 표상한 것의 원천으로 보았다. 베트남 관음상에 대한 중국문화의 영향은 그 영향이 직접적이며 장기간에 걸쳐 최근까지도 이어져져 왔기 때문에, 인도문화가 남긴 흔적들보다 더 분명하게 나타난다. 이러한 문화적 유사성으로 인하여 비평가들은 베트남 관음상을 중국 관음상의 단순한 복제품으로 간주하게 만들었다. 본 논문에서 저자는 베트남 관음상은 여러 외국의 특징들을 흡수해 왔지만, 동시에 베트남의 심리와 행동, 개념, 토착 종교적 믿음에 상응하는 그만의 독특한 특징을 반영했다는 사실을 주장할 것이다. 천개의 팔과 눈을 가진 베트남 관음상은 인도와 중국의 영향을 받았지만, 베트남 특유의 예술적 가치를 지니고 있으며, 아시아 관음상의 무한히 풍부한 다양성에 또 하나의 기여를 해 왔다.