• ETRI Journal
• /
• 제33권6호
• /
• pp.935-944
• /
• 2011

As a growing number of individuals are exposed to surveillance cameras, the need to prevent captured videos from being used inappropriately has increased. Privacy-related information can be protected through video encryption during transmission or storage, and several algorithms have been proposed for such purposes. However, the simple way of evaluating the security by counting the number of brute-force trials is not proper for measuring the security of video encryption algorithms, considering that attackers can devise specially crafted attacks for specific purposes by exploiting the characteristics of the target video codec. In this paper, we introduce a new attack for recovering contour information from encrypted H.264 video. The attack can thus be used to extract face outlines for the purpose of personal identification. We analyze the security of previous video encryption schemes against the proposed attack and show that the security of these schemes is lower than expected in terms of privacy protection. To enhance security, an advanced block shuffling method is proposed, an analysis of which shows that it is more secure than the previous method and can be an improvement against the proposed attack.

• 한국컴퓨터정보학회논문지
• /
• 제24권1호
• /
• pp.137-143
• /
• 2019

As the performance of computers increases, the use of deep learning, which has faced technical limitations in the past, is becoming more diverse. In many fields, deep learning has contributed to the creation of added value and used on the bases of more data as the application become more divers. The process for obtaining a better performance model will require a longer time than before, and therefore it will be necessary to find an optimal model that shows the best performance more quickly. In the artificial neural network modeling a tuning process that changes various elements of the neural network model is used to improve the model performance. Except Gride Search and Manual Search, which are widely used as tuning methods, most methodologies have been developed focusing on heuristic algorithms. The heuristic algorithm can get the results in a short time, but the results are likely to be the local optimal solution. Obtaining a global optimal solution eliminates the possibility of a local optimal solution. Although the Brute Force Method is commonly used to find the global optimal solution, it is not applicable because of an infinite number of hyper parameter combinations. In this paper, we use a statistical technique to reduce the number of possible cases, so that we can find the global optimal solution.

• Nuclear Engineering and Technology
• /
• 제56권4호
• /
• pp.1330-1338
• /
• 2024

Information on isotopic composition and geometric structure is necessary for identifying a true warhead. Nevertheless, such classified information should be protected physically or electronically. With a novel Hash encryption algorithm, this paper presents a Monte Carlo-based design of a neutron activation analysis verification module. The verification module employs a thermal neutron source, a non-uniform mask (physically encrypting information about isotopic composition and geometric structure), a gamma detector array, and a Hash encryption algorithm (for electronic encryption). In the physical field, a non-uniform mask is designed to distort the characteristic gamma rays emitted by the inspected item. Furthermore, as part of the Hash algorithm, a key is introduced to encrypt the data and improve the system resolution through electronic design. In order to quantify the difference between items, Hamming distance is used, which allows data encryption and analysis simultaneously. Simulated inspections of simple objects are used to quantify system performance. It is demonstrated that the method retains superior resolution even with 1% noise level. And the performances of anti-statistical attack and anti-brute force cracking are evaluated and found to be very excellent. The verification method lays a solid foundation for nuclear disarmament verification in the upcoming era.

• 한국콘텐츠학회논문지
• /
• 제17권7호
• /
• pp.571-578
• /
• 2017

사용자가 스마트 기기 및 개인 PC를 통해 정보 시스템에 접근하기 위해서는 사용자 인증을 통한 정당한 사용자임을 증명해야한다. 이때 사용되는 사용자 인증 기술에는 ID/PW 기반 인증기술, OTP(One-Time-Password), 인증서, 보안카드, 지문인식 등 다양하다. 그 중에서 ID/PW 기반 인증기술은 사용자에게 익숙하지만 무작위 대입공격, 키로깅, 사전공격 등의 패스워드 공격에 취약하며 이러한 공격에 대응하기 위해 사용자는 복잡한 패스워드 조합 규정에 따라 주기적으로 패스워드를 변경해야한다. 본 논문에서는 기존 ID/PW 기반 인증 기술보다 보안성을 높이면서 패스워드를 사용하지 않고 스마트폰의 USIM 정보를 이용한 사용자 인증 시스템을 설계 및 구현하였다.

• Journal of Multimedia Information System
• /
• 제8권3호
• /
• pp.147-158
• /
• 2021

Versatile Video Coding (VVC) is the latest video coding standard developed by Joint Video Exploration Team (JVET). In VVC, the quadtree plus multi-type tree (QT+MTT) structure of coding unit (CU) partition is adopted, and its computational complexity is considerably high due to the brute-force search for recursive rate-distortion (RD) optimization. In this paper, we aim to reduce the time complexity of inter-picture prediction mode since the inter prediction accounts for a large portion of the total encoding time. The problem can be defined as classifying the split mode of each CU. To classify the split mode effectively, a novel convolutional neural network (CNN) called multi-level tree (MLT-CNN) architecture is introduced. For boosting classification performance, we utilize additional information including inter-picture information while training the CNN. The overall algorithm including the MLT-CNN inference process is implemented on VVC Test Model (VTM) 11.0. The CUs of size 128×128 can be the inputs of the CNN. The sequences are encoded at the random access (RA) configuration with five QP values {22, 27, 32, 37, 42}. The experimental results show that the proposed algorithm can reduce the computational complexity by 11.53% on average, and 26.14% for the maximum with an average 1.01% of the increase in Bjøntegaard delta bit rate (BDBR). Especially, the proposed method shows higher performance on the sequences of the A and B classes, reducing 9.81%~26.14% of encoding time with 0.95%~3.28% of the BDBR increase.

• Nuclear Engineering and Technology
• /
• 제54권3호
• /
• pp.790-802
• /
• 2022

Large-scale reactor simulation often requires the use of Monte Carlo calculation techniques to estimate important reactor parameters. One drawback of these Monte Carlo calculation techniques is they inevitably result in some uncertainty in calculated quantities. The present study includes parametric uncertainty quantification (UQ) and sensitivity analysis (SA) on the Advanced Test Reactor Critical (ATRC) facility housed at Idaho National Laboratory (INL) and addresses some complications due to Monte Carlo uncertainty when performing these analyses. This approach for UQ/SA includes consideration of Monte Carlo code uncertainty in computed sensitivities, consideration of uncertainty from directly measured parameters and a comparison of results obtained from brute-force Monte Carlo UQ versus UQ obtained from a surrogate model. These methodologies are applied to the uncertainty and sensitivity of k_eff for two sets of uncertain parameters involving fuel plate geometry and fuel plate composition. Results indicate that the less computationally-expensive method for uncertainty quantification involving a linear surrogate model provides accurate estimations for k_eff uncertainty and the Monte Carlo uncertainty in calculated k_eff values can have a large effect on computed linear model parameters for parameters with low influence on k_eff.

• 정보보호학회논문지
• /
• 제32권6호
• /
• pp.1035-1043
• /
• 2022

사물인터넷(IoT), 클라우드 컴퓨팅, 빅데이터 등의 확산으로 애플리케이션에 대한 고속 암호화의 필요성이 대두되고 있다. GPU 최적화는 GPU가 이론적으로 얻은 암호 분석 결과 또는 축소된 버전을 합리적인 시간에 검증하는데 사용될 수 있다. 본 논문에서는 다양한 환경에서 구현되고 있는 PIPO 경량암호를 대상으로 GPU 상에서 구현하였다. PIPO에 대한 무차별 대입 공격을 고려하여 최적 구현하였다. 특히 비트 슬라이싱 기법을 적용한 최적화 구현과 GPU 요소를 최대한 사용하였다. 결과적으로 제안 기법의 구현은 RTX 3060 환경에서 초당 약 195억의 처리량을 보여 이전 연구 보다 약 122배 높은 처리량을 달성하였다.

• Nuclear Engineering and Technology
• /
• 제48권3호
• /
• pp.765-772
• /
• 2016

In analyzing residual radiation, researchers generally use a two-step Monte Carlo (MC) simulation. The first step (MC1) simulates neutron transport, and the second step (MC2) transports the decay photons emitted from the activated materials. In this process, the stochastic uncertainty estimated by the MC2 appears only as a final result, but it is underestimated because the stochastic error generated in MC1 cannot be directly included in MC2. Hence, estimating the true stochastic uncertainty requires quantifying the propagation degree of the stochastic error in MC1. The brute force technique is a straightforward method to estimate the true uncertainty. However, it is a costly method to obtain reliable results. Another method, called the adjoint-based method, can reduce the computational time needed to evaluate the true uncertainty; however, there are limitations. To address those limitations, we propose a new strategy to estimate uncertainty propagation without any additional calculations in two-step MC simulations. To verify the proposed method, we applied it to activation benchmark problems and compared the results with those of previous methods. The results show that the proposed method increases the applicability and user-friendliness preserving accuracy in quantifying uncertainty propagation. We expect that the proposed strategy will contribute to efficient and accurate two-step MC calculations.

• 한국농림기상학회지
• /
• 제16권4호
• /
• pp.304-308
• /
• 2014

벚꽃의 만개일은 관련 행사일정을 결정하는 중요한 요소로써 생육기간 중 기온에 따른 변화의 폭이 크다. 이를 예측하기 위한 방법으로는 벚꽃의 발달을 휴면기와 생장기의 2단계로 구분하여 저온(chill)과 고온(heat) 요구에 대한 온도시간(thermal time) 누적을 기술하는 모형이 개발되어 있다. 하지만 모수 추정시 모수공간내 일정 간격의 격자 전체를 계산하여 많은 시간을 소모한다는 단점이 있었다. 본 연구에서는 기존모형이 고려하지 않던 벚꽃 발달의 중간단계 관측자료를 활용하여 고온요구에 대한 새로운 조건을 추가하고, 이를 기반으로 각 모수를 순차적으로 추정하여 최적화 시간을 단축하는 새로운 방법을 제안한다. 미국 워싱턴 DC 지역의 벚꽃개화 관측 자료를 기준으로 검증한 결과, 기존 모형에서 제안된 모수와 근사한 값을 단축된 시간 내에 계산해내는 것을 확인하였다.

• 한국컴퓨터그래픽스학회논문지
• /
• 제29권3호
• /
• pp.127-135
• /
• 2023

조직병리에서 전체 슬라이드 영상의 정확한 분할은 질병 진단과 치료 계획에 매우 중요한 작업이다. 그러나 전체 슬라이드 영상은 크기가 크고 조직의 형태, 염색 및 촬영 조건이 다양하기 때문에 기존의 자동 영상 분할 알고리즘을 항상 적용하는 것은 어렵다. 최근 인간의 전문 지식과 알고리즘을 결합한 대화형 영상 분할 기술의 발전은 전체 슬라이드 영상 분할의 효율 성과 정확성을 개선할 수 있는 가능성을 보여주었다. 그러나 이러한 접근 방식은 동시에 어려운 과제를 제기하기도 했다. 본 논문에서는 다중 해상도 전체 슬라이드 영상을 활용하는 새로운 대화형 분할 방법인 ZoomISEG를 제안한다. 기존의 단일 스케일 방법과의 비교 및 ablation study를 통해 제안된 방법의 효율성과 성능을 입증한다. 실험 결과, 제안된 방법은 사람의 개입을 줄이면서도 최고 해상도 데이터를 사용하는 방식에 필적하는 정확도를 달성함을 확인했다.