딥블록: 웹 기반 딥러닝 교육용 플랫폼

조진성,김근모,고현민,김성민,김지섭,김봉재,Cho, Jinsung,Kim, Geunmo,Go, Hyunmin,Kim, Sungmin,Kim, Jisub,Kim, Bongjae 한국인터넷방송통신학회 2021 한국인터넷방송통신학회 논문지 Vol.21 No.3

최근 인공지능을 사용한 연구나 기업의 프로젝트가 활발하게 이루어지고 다양한 서비스나 시스템이 인공지능 기술과 접목되어 점점 더 지능화되고 있다. 이에 따라 인공지능의 기법 중 하나인 딥러닝에 대한 관심과 이를 학습하려는 사람들이 증가했다. 딥러닝을 학습하기 위해서는 딥러닝 이론 이외에도 컴퓨터 프로그래밍, 수식 등 많은 지식들이 요구된다. 이는 초심자에게 높은 진입장벽으로 작용한다. 따라서 본 연구에서는 초심자가 프로그래밍 및 수식 등을 고려하지 않고 DNN, CNN 등과 같은 딥러닝의 기본적인 모델을 구현할 수 있는 DeepBlock이라는 웹 기반 교육용 딥러닝 플랫폼을 설계 및 구현하였다. 제안한 DeepBlock을 이용하여 딥러닝에 관심을 가진 학생들이나 초심자들의 교육에 활용이 가능하다. Recently, researches and projects of companies based on artificial intelligence have been actively carried out. Various services and systems are being grafted with artificial intelligence technology. They become more intelligent. Accordingly, interest in deep learning, one of the techniques of artificial intelligence, and people who want to learn it have increased. In order to learn deep learning, deep learning theory with a lot of knowledge such as computer programming and mathematics is required. That is a high barrier to entry to beginners. Therefore, in this study, we designed and implemented a web-based deep learning platform called DeepBlock, which enables beginners to implement basic models of deep learning such as DNN and CNN without considering programming and mathematics. The proposed DeepBlock can be used for the education of students or beginners interested in deep learning.

음성 데이터 전처리 기법에 따른 뉴로모픽 아키텍처 기반 음성 인식 모델의 성능 분석

조진성,김봉재 한국인터넷방송통신학회 2022 한국인터넷방송통신학회 논문지 Vol.22 No.3

SNN (Spiking Neural Network) operating in neuromorphic architecture was created by mimicking human neural networks. Neuromorphic computing based on neuromorphic architecture requires relatively lower power than typical deep learning techniques based on GPUs. For this reason, research to support various artificial intelligence models using neuromorphic architecture is actively taking place. This paper conducted a performance analysis of the speech recognition model based on neuromorphic architecture according to the speech data preprocessing technique. As a result of the experiment, it showed up to 84% of speech recognition accuracy performance when preprocessing speech data using the Fourier transform. Therefore, it was confirmed that the speech recognition service based on the neuromorphic architecture can be effectively utilized. 뉴로모픽 아키텍처에서 동작하는 SNN (Spiking Neural Network) 은 인간의 신경망을 모방하여 만들어졌다. 뉴로모픽 아키텍처 기반의 뉴로모픽 컴퓨팅은 GPU를 이용한 딥러닝 기법보다 상대적으로 낮은 전력을 요구한다. 이와같은 이유로 뉴로모픽 아키텍처를 이용하여 다양한 인공지능 모델을 지원하고자 하는 연구가 활발히 일어나고 있다. 본논문에서는 음성 데이터 전처리 기법에 따른 뉴로모픽 아키텍처 기반의 음성 인식 모델의 성능 분석을 진행하였다. 실험결과 푸리에 변환 기반 음성 데이터 전처리시 최대 84% 정도의 인식 정확도 성능을 보임을 확인하였다. 따라서 뉴로모픽 아키텍처 기반의 음성 인식 서비스가 효과적으로 활용될 수 있음을 확인하였다.

WBAN을 위한 MAC 프로토콜 기술 동향

조진성,김윤희,이민수 한국컴퓨터통신연구회 2008 OSIA Standards & Technology Review Vol.33 No.3

Diagnostic modality in field

조진성 대한응급의학회 2022 대한응급의학회 학술대회초록집 Vol.2022 No.2

MANAGEMENT OF HYPOCALCEMIA

조진성 대한갑상선-내분비외과학회 2015 대한갑상선-내분비외과학회 학술대회 논문집 Vol.- No.-

T1 T2 갑상선 유두암에서 도약 측경부 전이의 위험인자 및 치료전략

조진성,윤현철,김병천,신선형,송영주,박민호,윤정한,제갈영종 대한외과학회 2010 대한외과학회 학술대회 초록집 Vol.2010 No.11

복부 대동맥류의 임상적 고찰

조진성,최수진나 대한외과학회 2003 Annals of Surgical Treatment and Research(ASRT) Vol.65 No.6

Purpose: In patients with small or large aneurysms, the decision for surgical treatment is not so simple. The mortality of ruptured abdominal aortic aneurysm (AAA) is high. This study was designed to retrospectively analyse the clinical characteristics of patients with AAA. Methods: Ninety-one cases of AAA were surgically treated between January 1991 and January 2003 at the Department of Surgery, Chonnam National University Hospital. Patients were divided into 49 elective cases and 42 emergency cases, and retrospective analysed on the basis of age, sex, chief complaints, physical examination, associated diseases, size of aneurysm, diagnostic modalities, operative mortality and causes of death. Results: The initial presentations were mainly palpable masses in the elective cases. On the other hand, in the emergency cases which were ruptured, many patients complained of abdominal or back pain. There was a positive relationship between the size of AAA and the incidence of the rupture in our study, especially in the case of transverse diameters above 10 cm (P<0.001). There was no death in the elective cases, but there were 22 surgical mortalities in the 42 emergency cases (52.3%, P<0.001). Overall surgical mortality was 24.1%. The causes of death were intraoperative and postoperative bleeding (11), myocardial infarction (5), acute renal failure (4), and sepsis (2). Conclusion: Surgical mortality in ruptured AAA was high. Consequently, surgical intervention is recommended and the operation must be performed. In that way we can reduce the operative mortality and improve the treatment outcome. (J Korean Surg Soc 2003;65:554-558)

국내 헬리콥터 응급의료체계의 도입과 현황

조진성,양혁준 대한의사협회 2020 대한의사협회지 Vol.63 No.4

As most medical facilities in Korea have been concentrated in large cities, the need to improve emergency medical services in islands and mountainous areas has emerged. Consequently, the Ministry of Health and Welfare and local governments have introduced emergency medical helicopters (known as helicopter ambulances or air ambulances) with doctors in medically vulnerable areas. Having been introduced in two regions in 2011, air ambulances are operational in seven regions as of the end of 2019. The flight time is from sunrise to sunset, except in Gyeonggi province, which is open all day. Although the criteria for transport vary depending on whether an ambulance is available for operation, it is basically intended for emergency critical diseases, such as severe trauma, stroke, and acute myocardial infarction. From September 23, 2011 to December 31, 2018, a total of 10,367 transfer requests were received, which included 534 (5.2%) interruptions, 2,657 (25.6%) rejects, and 7,176 (69.2%) transfers. A total of 7,209 patients were transferred during this period, which included 1,693 (23.5%) patients of severe trauma, 1,149 (15.9%) patients of stroke, 802 (11.1%) patients of acute myocardial infarction, and 3,565 (49.5%) patients suffering from other emergency diseases. Some economic research on air ambulances in Korea has been reported to be cost-effective, but additional research should be performed. In the future, it is necessary to widen the area of operation of air ambulances and find alternative means of transporting patients during unfavorable conditions such as night or bad weather.

cdma2000 무선 패킷 데이터 서비스의 이동성 관리

조진성 한국정보과학회 정보통신 소사이어티 2003 정보통신 기술 Vol.17 No.2

NAAL: 뉴로모픽 아키텍처 추상화 기반 이기종 IoT 기기 제어용 소프트웨어

조진성,김봉재 (사)한국스마트미디어학회 2022 스마트미디어저널 Vol.11 No.3

Neuromorphic computing generally shows significantly better power, area, and speed performance than neural network computation using CPU and GPU. These characteristics are suitable for resource-constrained IoT environments where energy consumption is important. However, there is a problem in that it is necessary to modify the source code for environment setting and application operation according to heterogeneous IoT devices that support neuromorphic computing. To solve these problems, NAAL was proposed and implemented in this paper. NAAL provides functions necessary for IoT device control and neuromorphic architecture abstraction and inference model operation in various heterogeneous IoT device environments based on common APIs of NAAL. NAAL has the advantage of enabling additional support for new heterogeneous IoT devices and neuromorphic architectures and computing devices in the future. 뉴로모픽 컴퓨팅은 일반적으로 CPU와 GPU를 이용하여 신경망 연산을 하는 것보다 전력, 면적, 속도 측면에서 매우 뛰어난 성능을 보여준다. 이러한 특성은 에너지 사용량이 중요시되는 자원 제약적인 IoT 환경에 매우 적합하다. 하지만 뉴로모픽 컴퓨팅을 지원하는 이기종 IoT 기기에 따라 환경설정 및 응용 프로그램 동작을 위한 소스코드의 수정이 필요하다는 문제점을 가지고 있다. 이러한 문제점을 해결하고자 본 논문에서는 NAAL을 제안하고 구현하였다. NAAL은 공통의 API를 기반으로 다양한 이기종 IoT 기기 환경에서 IoT 기기의 제어와 뉴로모픽 아키텍처의 추상화 및 추론 모델 동작에 필요한 기능을 제공한다. NAAL은 향후 새로운 이기종 IoT 기기 및 뉴로모픽 아키텍처와 컴퓨팅 장치의 추가적인 지원이 가능하다는 장점을 가진다.