연구성과로 돌아가기
2020 연구성과 (270 / 270)
※ 컨트롤 + 클릭으로 열별 다중 정렬 가능합니다.
Excel 다운로드
| WoS | SCOPUS | Document Type | Document Title | Abstract | Authors | Affiliation | ResearcherID (WoS) | AuthorsID (SCOPUS) | Author Email(s) | Journal Name | JCR Abbreviation | ISSN | eISSN | Volume | Issue | WoS Edition | WoS Category | JCR Year | IF | JCR (%) | FWCI | FWCI Update Date | WoS Citation | SCOPUS Citation | Keywords (WoS) | KeywordsPlus (WoS) | Keywords (SCOPUS) | KeywordsPlus (SCOPUS) | Language | Publication Stage | Publication Year | Publication Date | DOI | JCR Link | DOI Link | WOS Link | SCOPUS Link |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| ○ | Conference paper | Waveform-based End-to-end Deep Convolutional Neural Network with Multi-scale Sliding Windows for Weakly Labeled Sound Event Detection | In this paper, a waveform-based end-to-end sound event detection algorithm that detects and classifies sound events using a deep convolutional neural network architecture is proposed. While most machine-learning-based acoustic signal processing systems utilize hand-crafted feature vectors e.g. log-Mel spectrogram, end-to-end methods, which utilize raw input data, have recently been investigated for use in various applications. Therefore, we develop an end-to-end architecture for sound event detection tasks with convolutional neural networks. The proposed model consists of multi-scale time frames and networks that handle both short and long signal characteristics; the frame slides by 0.1 second to provide a sufficiently fine resolution. The element network for each time frame consists of several one-dimensional convolutional neural networks with a deeply stacked structure. The results of the element networks are averaged and gated by sound activity detection. In order to handle unlabeled data, the trained networks are enhanced using the mean-teacher model. A decision is made via double thresholding, and the results are enhanced using class-wise minimum gap/length compensation. To evaluate our proposed approach, simulations are performed with development data from DCASE 2019 Task 4, and the results show that the proposed algorithm had a macro-averaged F1 score of 31.7% for the DCASE 2019 development dataset, 30.2% for the DCASE 2018 evaluation dataset, and 26.7% for the DCASE 2019 evaluation dataset. © 2020 IEEE. | Lee, Seokjin; Kim, Minhan | School of Electronics Engineering, Kyungpook National University, Daegu, South Korea; School of Electronics Engineering, Kyungpook National University, Daegu, South Korea | 36174416200; 57216617123 | 2020 International Conference on Artificial Intelligence in Information and Communication, ICAIIC 2020 | 0.45 | 2025-06-25 | 5 | convolutional neural network; end-to-end; sound event detection; waveform; weakly supervised | Convolution; Deep neural networks; Network architecture; Signal processing; Acoustic signals; Activity detection; Double thresholding; Feature vectors; Fine resolution; Signal characteristic; Sound event detection; Stacked structure; Convolutional neural networks | English | Final | 2020 | 10.1109/icaiic48513.2020.9064985 | 바로가기 | 바로가기 | |||||||||||||||||||
| ○ | ○ | Proceedings Paper | Weakly-supervised US breast tumor characterization and localization with a box convolution network | In US breast tumor diagnosis, machine learning approaches for the malignancy classification and the mass localization have been attracting many researchers to improve the diagnostic sensitivity and specificity while reducing the image interpretation time. Recently, fully-supervised deep learning methods showed their promising results in those tasks. However, the full supervision for the localization requires human efforts and time to annotate ground truth regions. In this paper, we present a weakly-supervised deep network which can localize breast masses in US images from only diagnostic labels (i.e., malignant and benign). Specifically, we exploit a flexible convolution method, which learns the size and offset of the convolution kernel, in the classification network to detect more relevant regions of breast masses against their various size and shape. Experimental results show that the proposed network outperform conventional CNN models, such as VGG-16 and VGG-16 with dilated convolution. The proposed model achieved 89.03% in the binary classification accuracy. To evaluate the localization performance with weakly-supervised manners, we also compared class activation maps for each instance with manual masks of breast mass in terms of the Dice similarity coefficient and localization recall. The experimental results also demonstrate that the deep network with the adjustable convolution layers can clinically relevant features of breast mass and its surrounding area for both benign and malignant cases. | Kim, Chanho; Kim, Won Hwa; Kim, Hye Jung; Kim, Jaeil | Kyungpook Natl Univ, Sch Comp Sci & Engn, Daegu, South Korea; Kyungpook Natl Univ, Dept Radiol, Chilgok Hosp, Daegu, South Korea; Kyungpook Natl Univ, Dept Radiol, Sch Med, Daegu, South Korea | 57216946967; 36081886500; 57203506201; 57211615348 | threeyears@gmail.com; | MEDICAL IMAGING 2020: COMPUTER-AIDED DIAGNOSIS | 0277-786X | 1996-756X | 11314 | 1.93 | 2025-06-25 | 3 | 6 | Breast Cancer; Covolutional Neural Networks; Tumor Classification; Tumor Localization; Ultrasound Imaging; Weakly-supervised Learning | ULTRASOUND; LESIONS | breast cancer; covolutional neural networks; tumor classication; tumor localization; ultrasound imaging; weakly-supervised learning | Convolution; Deep learning; Image enhancement; Learning systems; Medical imaging; Tumors; Binary classification; Classification networks; Convolution methods; Image interpretation; Localization performance; Machine learning approaches; Sensitivity and specificity; Similarity coefficients; Computer aided diagnosis | English | 2020 | 2020 | 10.1117/12.2549203 | 바로가기 | 바로가기 | 바로가기 | ||||||||||
| ○ | Conference paper | Weight dropout for preventing neural networks from overfitting | This paper briefly introduces an enhanced neural network regularization method, so called weight dropout, in order to prevent deep neural networks from overfitting. In suggested method, the fully connected layer jointly used with weight dropout is a collection of layers in which the weights between nodes are dropped randomly on the process of training. To accomplish the desired regularization method, we propose a building blocks with our weight dropout mask and CNN. The performance of proposed method has been compared with other previous methods in the domain of image classification and segmentation for the evaluation purpose. The results show that the proposed method gives successful performance accuracies in several datasets. © 2020 IEEE. | Sanjar, Karshiev; Rehman, Abdul; Paul, Anand; Jeonghong, Kim | Kyungpook National University, Department of Computer Science, South Korea; Kyungpook National University, Department of Computer Science, South Korea; Kyungpook National University, Department of Computer Science, South Korea; Kyungpook National University, Department of Computer Science, South Korea | 57210910507; 57200894071; 56650522400; 55138548100 | jhk@knu.ac.kr; | 2020 8th International Conference on Orange Technology, ICOT 2020 | 2.35 | 2025-06-25 | 19 | Component; Image classification; Overfitting; Regularization; Semantic segmentation; Weight dropout | Citrus fruits; Deep neural networks; Image segmentation; Building blockes; Overfitting; Regularization methods; Neural networks | English | Final | 2020 | 10.1109/icot51877.2020.9468799 | 바로가기 | 바로가기 | ||||||||||||||||||
| ○ | Editorial | Welcome message from conference chair | [No abstract available] | Jung, Soon Ki | Head of Graduate School of Computer Science and Engineering, Kyungpook National University, South Korea | 57226791905 | 2020 8th International Conference on Orange Technology, ICOT 2020 | 0 | 2025-06-25 | 0 | English | Final | 2020 | 10.1109/icot51877.2020.9468774 | 바로가기 | 바로가기 | |||||||||||||||||||||
| ○ | Editorial | Welcome messages | [No abstract available] | Jo, Myung-Hee | Kyungpook National University, South Korea | 55348712900 | 40th Asian Conference on Remote Sensing, ACRS 2019: Progress of Remote Sensing Technology for Smart Future | 0 | 2025-06-25 | 0 | English | Final | 2020 | 바로가기 | |||||||||||||||||||||||
| ○ | ○ | Review | Why most patients do not exhibit obstructive sleep apnea after mandibular setback surgery? | Maxillomandibular advancement (MMA) is effective for the treatment of obstructive sleep apnea (OSA). In previous studies, the airway was increased in the anteroposterior and transverse dimensions after MMA. However, the effect of the opposite of mandibular movement (mandibular setback) on the airway is still controversial. Mandibular setback surgery has been suggested to be one of the risk factors in the development of sleep apnea. Previous studies have found that mandibular setback surgery could reduce the total airway volume and posterior airway space significantly in both the one-jaw and two-jaw surgery groups. However, a direct cause-and-effect relationship between the mandibular setback and development of sleep apnea has not been clearly established. Moreover, there are only a few reported cases of postoperative OSA development after mandibular setback surgery. These findings may be attributed to a fundamental difference in demographic variables such as age, sex, and body mass index (BMI) between patients with mandibular prognathism and patients with OSA. Another possibility is that the site of obstruction or pattern of obstruction may be different between the awake and sleep status in patients with OSA and mandibular prognathism. In a case-controlled study, information including the BMI and other presurgical conditions potentially related to OSA should be considered when evaluating the airway. In conclusion, the preoperative evaluation and management of co-morbid conditions would be essential for the prevention of OSA after mandibular setback surgery despite its low incidence. | Kim, Jin-Wook; Kwon, Tae-Geon | Kyungpook Natl Univ, Dept Oral & Maxillofacial Surg, Sch Dent, 2177 Dalgubeol Daero, Daegu 41940, South Korea | 55862646000; 35205433300 | kwondk@knu.ac.kr; | MAXILLOFACIAL PLASTIC AND RECONSTRUCTIVE SURGERY | MAX PLAST RECONSTR S | 2288-8101 | 2288-8586 | 42 | 1 | ESCI | DENTISTRY, ORAL SURGERY & MEDICINE | 2020 | N/A | 0.43 | 2025-06-25 | 8 | 9 | Mandibular setback; Obstructive sleep apnea; Airway; Prognathism | CLASS-III PATIENTS; UPPER-AIRWAY CHANGES; MAXILLOMANDIBULAR ADVANCEMENT; PHARYNGEAL AIRWAY; COMPUTED-TOMOGRAPHY; BIMAXILLARY SURGERY; 3-DIMENSIONAL CHANGES; ORTHOGNATHIC SURGERY; SURROUNDING STRUCTURES; VOLUMETRIC CHANGES | Airway; Mandibular setback; Obstructive sleep apnea; Prognathism | airway obstruction; apnea hypopnea index; body mass; causal reasoning; clinical examination; comparative study; face surgery; human; mandible fracture; mandibular setback surgery; obesity; orthognathic surgery; polysomnography; positive end expiratory pressure; postoperative complication; preoperative evaluation; priority journal; prognathia; Review; risk factor; sleep disorder; sleep disordered breathing; upper respiratory tract obstruction | English | 2020 | 2020-03-17 | 10.1186/s40902-020-00250-x | 바로가기 | 바로가기 | 바로가기 | 바로가기 | |||
| ○ | ○ | Proceedings Paper | WIRE: An Automated Report Generation System using Topical and Temporal Summarization | The demand for a tool for summarizing emerging topics is increasing in modern life since the tool can deliver well-organized information to its users. Even though there are already a number of successful search systems, the system which automatically summarizes and organizes the content of emerging topics is still in its infancy. To fulfill such demand, we introduce an automated report generation system that generates a well-summarized human-readable report for emerging topics. In this report generation system, emerging topics are automatically discovered by a topic model and news articles are indexed by the discovered topics. Then, a topical summary and a timeline summary for each topic is generated by a topical multi-document summarizer and a timeline summarizer respectively. In order to enhance the apprehensibility of the users, the proposed report system provides two report modes. One is Today's Briefing which summarizes five discovered topics of every day, and the other is Full Report which shows a long-term view of each topic with a detailed topical summary and an important event timeline. | Noh, Yunseok; Shin, Yongmin; Park, Junmo; Kim, A-Yeong; Choi, Su Jeong; Song, Hyun-Je; Park, Seong-Bae; Park, Seyoung | Kyungpook Natl Univ, Daegu, South Korea; KT, Inst Convergence Technol, Seoul, South Korea; Jeonbuk Natl Univ, Jeonju, South Korea; Kyung Hee Univ, Seoul, South Korea | 54403595500; 57218706584; 57218705387; 42661508900; 56124323200; 35175084000; 7501838676; 14045781800 | ysnoh@sejong.knu.ac.kr;ymshin@sejong.knu.ac.kr;jmpark@sejong.knu.ac.kr;aykim@sejong.knu.ac.kr;sujeong.choi@kt.com;hyunje.song@jbnu.ac.kr;sbpark71@khu.ac.kr;seyoung@knu.ac.kr; | PROCEEDINGS OF THE 43RD INTERNATIONAL ACM SIGIR CONFERENCE ON RESEARCH AND DEVELOPMENT IN INFORMATION RETRIEVAL (SIGIR '20) | 0.81 | 2025-06-25 | 6 | 10 | report generation; text summarization; timeline summarization; topic discovery; text retrieval; image retrieval; deep neural networks | deep neural networks; image retrieval; report generation; text retrieval; text summarization; timeline summarization; topic discovery | Emerging topics; Human-readable; Multi-document; News articles; Report generation; Successful search; Topic Modeling; Information retrieval | English | 2020 | 2020 | 10.1145/3397271.3401409 | 바로가기 | 바로가기 | 바로가기 | ||||||||||||||
| ○ | Article | Yoo Chijin's Strategy to Popularize Singeuk in Colonial Korea: The Story of Chunhyang and Porgy | Dorothy and DuBose Heyward's Porgy was performed in colonial Korea in 1937. This paper explains the process of performing Porgy, in connection with the strategy to popularize the singeuk (new drama) by Yoo Chijin, a playwright and director who argued that one could popularize singeuk by promoting performance of colonial Korean dramas. Yoo's strategy encountered great difficulties because of opposition by members of the Research Association of Theatrical Art and strict Japanese censorship. Yoo tried to overcome these problems pertaining to realistic plays such as Slums and The Cow; however, his efforts were in vain. Therefore, Yoo undertook a new strategy of adapting the traditional The Story of Chunhyang, a play of "romanticism based on realism," which bypassed censorship by expressing the reality of the era metaphorically and amused the audience of the grand theatre with songs and dances. Porgy's songs and dances influenced The Story of Chunhyang. Although the audience responded favorably to The Story of Chunhyang, critics found fault with the fact that Yoo was a playwright who followed the practices of commercial theatre. Yoo tried to refute their criticism by producing Porgy, a performance of "romanticism on the basis of realism." He argued that The Story of Chunhyang reflected the latest theatre trends in the United States. | Jaesuk, Kim | Kyungpook Natl Univ, Dept Korean Language & Literature, Daegu, South Korea | ASIAN THEATRE JOURNAL | ASIAN THEATRE J | 0742-5457 | 1527-2109 | 37 | 2 | AHCI | ASIAN STUDIES;THEATER | 2020 | N/A | 0 | English | 2020 | 2020-가을 | 10.1353/atj.2020.0035 | 바로가기 | 바로가기 | 바로가기 |
페이지 이동: