• 대한임베디드공학회논문지
• /
• 제15권2호
• /
• pp.43-49
• /
• 2020

In this paper, we propose a stage-wise knowledge transfer method that uses block-wise retraining to transfer the useful knowledge of a pre-trained residual network (ResNet) in a teacher-student framework (TSF). First, multiple hint information transfer and block-wise supervised retraining of the information was alternatively performed between teacher and student ResNet models. Next, Softened output information-based knowledge transfer was additionally considered in the TSF. The results experimentally showed that the proposed method using multiple hint-based bottom-up knowledge transfer coupled with incremental block-wise retraining provided the improved student ResNet with higher accuracy than existing KD and hint-based knowledge transfer methods considered in this study.

• International Journal of Internet, Broadcasting and Communication
• /
• 제15권3호
• /
• pp.231-238
• /
• 2023

As the demand for distance education increases, interest in the management of learners' rights is increasing. Blockchain technology is a technology that guarantees the integrity of the learner's learning history, and enables learner-led learning control, data security, and sharing of learning resources. In this paper, we proposed a blockchain technology-based learning management system based on Hyperledger Fabric that can be verified through permission between nodes among blockchain platforms. Learning resources can be shared differentially according to the learning progress. Also the percentage of individual learners that can be managed. As a result of the study, the superiority of the platform in terms of convenience compared to the existing platform was demonstrated. As a result of the performance evaluation for the research in this paper, it was confirmed that the convenience was improved by more than 5%, and the performance was 4-5% superior to the existing platform in terms of learner satisfaction.

• 한국빅데이터학회지
• /
• 제7권1호
• /
• pp.1-14
• /
• 2022

대부분의 블록체인이 사용하는 작업증명 합의 알고리즘은 채굴이라는 형태로 대규모의 컴퓨팅리소스 낭비를 초래하고 있다. 작업증명의 컴퓨팅리소스 낭비를 줄이기 위해 유용한 작업증명 합의 알고리즘이 연구 되었으나 여전히 블록 생성 시 리소스 낭비와 채굴의 중앙화 문제가 존재한다. 본 논문에서는 블록생성을 위한 상대적으로 비효율적인 연산 과정을 분산 인공지능 모델 학습으로 대체하여 블록생성 시 리소스 낭비문제를 해결하였다. 또한 학습 과정에 참여한 노드들에게 공평한 보상을 제공함으로써 컴퓨팅파워가 약한 노드의 참여 동기를 부여했고, 기존 중앙 집중 인공지능 학습 방식에 근사한 성능은 유지하였다. 제안된 방법론의 타당성을 보이기 위해 분산 인공지능 학습이 가능한 블록체인 네트워크를 구현하여 리소스 검증을 통한 보상 분배를 실험 하였고, 기존 중앙 학습 방식과 블록체인 분산 인공지능 학습 방식의 결과를 비교하였다. 또한 향후 연구로 블록체인 메인넷과 인공지능 모델 확장 시 발생 할 수 있는 문제점과 개발 방향성을 제시함으로서 논문을 마무리 하였다.

• Journal of Platform Technology
• /
• 제11권1호
• /
• pp.72-84
• /
• 2023

With the rise of the Internet of Things, the security of such lightweight computing environments has become a hot topic. Lightweight block ciphers that can provide efficient performance and security by having a relatively simpler structure and smaller key and block sizes are drawing attention. Due to these characteristics, they can become a target for new attack techniques. One of the new cryptanalytic attacks that have been attracting interest is Neural cryptanalysis, which is a cryptanalytic technique based on neural networks. It showed interesting results with better results than the conventional cryptanalysis method without a great amount of time and cryptographic knowledge. The first work that showed good results was carried out by Aron Gohr in CRYPTO'19, the attack was conducted on the lightweight block cipher SPECK-/32/64 and showed better results than conventional differential cryptanalysis. In this paper, we first apply the Differential Neural Distinguisher proposed by Aron Gohr to the block ciphers HIGHT and GOST to test the applicability of the attack to ciphers with different structures. The performance of the Differential Neural Distinguisher is then analyzed by replacing the neural network attack model with five different models (Multi-Layer Perceptron, AlexNet, ResNext, SE-ResNet, SE-ResNext). We then propose a Related-key Neural Distinguisher and apply it to the SPECK-/32/64, HIGHT, and GOST block ciphers. The proposed Related-key Neural Distinguisher was constructed using the relationship between keys, and this made it possible to distinguish more rounds than the differential distinguisher.

• 지구물리와물리탐사
• /
• 제25권2호
• /
• pp.59-70
• /
• 2022

최근의 연구들을 통해 기계학습은 탄성파 해석 분야에 그 적용 범위를 확장하고 있으며, 탄성파 해석에서 중요한 탄성파 층서 구분을 수행하는 합성곱 신경망들의 개발도 수행되었다. 하지만 지도 학습의 경우 대량의 학습 자료가 필요하며, 비용과 시간의 한계로 탄성파 층서구분의 지도학습은 학습 자료의 부족이 문제가 될 수 있다. 이번 연구에서는 자료 부족 문제를 보완하기위해 탄성파 단면에 패치 분할과 자료증강을 적용하였다. 또한 패치 분할로 손실될 수 있는 공간정보를 제공하기 위해 깊이를 고려할 수 있는 인공 채널을 생성하여 추가하였다. 실험을 위한 학습 모델로 U-Net을 사용하였으며, 층서 구분을 위한 학습 자료가 제공되는 F3 block 자료를 이용하여 학습과 예측 결과에 대한 평가를 수행하였다. 분석 결과 자료증강과 인공 채널의 추가로 패치 기반의 층서 구분 학습 모델을 개선할 수 있음을 확인하였다.

• International journal of advanced smart convergence
• /
• 제5권2호
• /
• pp.66-72
• /
• 2016

With the growing shift from traditional educational approaches and studying to the more digital classroom, using electronic textbooks and digital native's demand, there is a growing need to develop new methods for learn Kanji characters for children. The purpose of this study is to help children learn the basic Kanji by using stop motion and block methods, and approaching the basic Kanji character education with a more innovative and interactive smart phone APP. In the development of this smart phone App for children's Kanji character education proposed in this study, 100 basic Kanji characters for children are selected. These 100 characters are required for the stop motion animation production, where each selected Kanji is created as a stop-motion animation utilizing a variety of techniques, such as storytelling, to better engage children. The intelligent App is designed with image recognition technology, so that in the learning process children take a picture for the assembled block using their smart phone, the APP then recognizes whether it is assembled correctly, and then plays an animation corresponding to the assembled Kanji character.

• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 2000년도 추계종합학술대회 논문집(4)
• /
• pp.195-198
• /
• 2000

Fractal image compression can reduce the size of image data by contractive mapping of original image. The mapping is affine transformation to find the block(called range block) which is the most similar to the original image. Fractal is very efficient way to reduce the data size. However, it has high distortion rate and requires long encoding time. In this paper, we present the simulation result of fractal and VQ hybrid systems which use different clustering algorithms, normal and improved competitive learning SOFM. The simulation results showed that the VQ hybrid fractal using improved competitive learning SOFM has better distortion rate than the VQ hybrid fractal using normal SOFM.

• 한국지능시스템학회:학술대회논문집
• /
• 한국퍼지및지능시스템학회 2000년도 추계학술대회 학술발표 논문집
• /
• pp.81-84
• /
• 2000

Fractal image compression can reduce the size of image data by contractive mapping of original image. The mapping is affine transformation to find the block(called range block) which is the most similar to the original image. Fractal is very efficient way to reduce the data size. However, it has high distortion rate and requires long encoding time. In this paper, we present the simulation result of fractal and VQ hybrid systems which use different clustering algorithms, normal and improved competitive learning SOFM. The simulation results showed that the VQ hybrid fractal using improved competitive learning SOFM has better distortion rate than the VQ hybrid fractal using normal SOFM.

• 한국통신학회논문지
• /
• 제26권9B호
• /
• pp.1292-1298
• /
• 2001

본 논문에서는 공간 영역에서의 블록 분류 (block classification)와 순방향 신경망 필터(feedforward neural network filter)를 이용한 블록 기반 부호화에서의 적응적 블록화 현상 제거 알고리듬을 제안하였다. 제안한 방법에서는 각 블록 경계를 인접 블록간의 통계적 특성을 이용하여 평탄 영역과 에지 영역으로 분류한 후, 각 영역에 대하여 블록화 현상이 발생하였다고 분류된 클래스에 대하여 적응적인 블록간 필터링을 수행한다. 즉, 평탄 영역으로 분류된 영역 중 블록화 현상이 발생한 영역은 오류 역전파 학습 알고리듬 (error backpropagation learning algorithm)에 의하여 학습된 2계층 (2-layer) 신경망 필터를 이용하여 블록화 현상을 제거하고, 복잡한 영역으로 분류된 영역 중 블록화 현상이 발생한 영역은 에지 성분을 보존하기 위하여 선형 내삽을 이용하여 블록간 인접 화소의 밝기 값만을 조정함으로써 블록화 현상을 제거한다. 모의 실험 결과를 통하여 제안한 방법이 객관적 화질 및 주관적 화질 측면에서 기존의 방법보다 그 성능이 우수함을 확인하였다.

• 한국지능시스템학회:학술대회논문집
• /
• 한국퍼지및지능시스템학회 1998년도 추계학술대회 학술발표 논문집
• /
• pp.560-567
• /
• 1998

This paper concerned about a imputed color printed material image in camera to decrease noise and distortion by processing median filtering with input image to identical condition. Also this paper proposed the way of compares a normal printed material with an abnormal printed material color tone with trained a learning of the error back-propagation to block classification by extracting five place from identical block(3${\times}$3) of color printed material R, G, B value. As a representative algorithm of multi-layer perceptron the error Back-propagation technique used to solve complex problems. However, the Error Back-propagation is algorithm which basically used a gradient descent method which can be converged to local minimum and the Back Propagation train include problems, and that may converge in a local minimum rather than get a global minimum. The network structure appropriate for a given problem. In this paper, a good result is obtained by improve initial condition and adjust th number of hidden layer to solve the problem of real time process, learning and train.