• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.12 no.6
• /
• pp.135-139
• /
• 2012

The wide diffusion of internet and the overall increased reliance on technology for information communication, dissemination and gathering have created an unparalleled mass of data. Sifting through this data is defining and will define in the foreseeable future a big part of contemporary computer science. Within this data, a growing proportion is given by personal information, which represents a unique opportunity to study human activities extensively and live. One important recurring challenge in many disciplines is the problem of people re-identification. In its broadest definition, re-identification is the problem of newly recognizing previously identified people, such as following an unknown person while he walks through many different surveillance cameras in different locations. Our goals is to review how several diverse disciplines define and meet this challenge, from person re-identification in video-surveillance to authorship attribution in text samples to distinguishing users based on their preferences of pictures. We further envision a situation where multidisciplinary solutions might be beneficial.

• Journal of the Korea Society of Computer and Information
• /
• v.26 no.3
• /
• pp.35-42
• /
• 2021

In this paper, we propose a deep learning-based person re-identification method using a three-dimensional RGB-Depth Xtion2 camera considering joint coordinates and dynamic features(velocity, acceleration). The main idea of the proposed identification methodology is to easily extract gait data such as joint coordinates, dynamic features with an RGB-D camera and automatically identify gait patterns through a self-designed one-dimensional convolutional neural network classifier(1D-ConvNet). The accuracy was measured based on the F1 Score, and the influence was measured by comparing the accuracy with the classifier model (JC) that did not consider dynamic characteristics. As a result, our proposed classifier model in the case of considering the dynamic characteristics(JCSpeed) showed about 8% higher F1-Score than JC.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2019.06a
• /
• pp.120-122
• /
• 2019

본 논문은 보행자 재 검출 알고리즘, 즉 person Re-Identification 알고리즘에 대하여 다루고 있다. 기존의 CNN 네트워크를 이용한 보행자 재 검출 알고리즘의 경우, 실제 감시 카메라 네트워크를 이용하여 보행자 재 검출을 할 경우 주변 환경 조건이 급격하게 변하는 경우 잘못 검출하는 경우가 발생하는 것을 확인할 수 있다. 이는 보행자 검출 후 해당 영역에 대하여 보행자 재 검출을 하는데 있어서 배경 부분의 변화에 영향을 받는다는 것을 의미한다. 따라서 본 논문에서는 배경 부분의 영향에 의한 효과를 줄이기 위하여, 보행자 영역 분리 알고리즘을 이용하여 보행자 영역을 분리한 후, 보행자 재 검출을 수행하는 연구를 진행한다.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2018.06a
• /
• pp.102-104
• /
• 2018

본 논문은 보행자 재 검출 알고리즘, 즉 person Re-Identification 알고리즘에 대하여 다루고 있다. 기존의 CNN 네트워크를 이용한 보행자 재 검출 알고리즘의 경우, 실제 감시 카메라 네트워크를 이용하여 보행자 재 검출을 할 경우 주변 환경 조건이 급격하게 변하는 경우 잘못 검출하는 경우가 발생하는 것을 확인할 수 있다. 이는 보행자 검출 후 해당 영역에 대하여 보행자 재 검출을 하는데 있어서 배경 부분의 변화에 영향을 받는다는 것을 의미한다. 따라서 본 논문에서는 배경 부분의 영향에 의한 효과를 줄이기 위하여, 보행자 영역 분리 알고리즘을 이용하여 보행자 영역을 분리한 후, 보행자 재 검출을 수행하는 연구를 진행한다.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2020.07a
• /
• pp.261-262
• /
• 2020

보행자 재 검출 알고리즘, 즉 person Re-Identification 알고리즘은 주어진 영상 내에 존재하는 보행자들 중 특정 보행자를 검출해내는 방법이다. 최근까지 보행자 재 검출 알고리즘에 대한 여러 연구가 진행되어 오고 있지만 기존의 CNN 네트워크를 이용한 보행자 재 검출 알고리즘의 경우, 실제 영상 데이터를 이용하여 보행자 재 검출을 할 경우 주변 환경 조건이나 조명의 조건, 보행자를 촬영한 방향 등에 따라 정확도가 떨어지는 현상이 발행한다. 이에 따라, 보행자 재 검출 알고리즘을 수행하는데 있어서, 조명 등의 조건에 구애 받지 않고 정확한 검출을 할 수 있도록 style transfer를 이용하여 영상을 변형하여 보행자 재검출을 수행하는 연구를 진행한다.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.15 no.12
• /
• pp.4456-4475
• /
• 2021

The system for re-identifying persons is used to find and verify the persons crossing through different spots using various cameras. Much research has been done to re-identify the person by utilising features with deep-learned or hand-crafted information. Deep learning techniques segregate and analyse the features of their layers in various forms, and the output is complex feature vectors. This paper proposes a distinctive framework called Integrated Level Feature Pattern (ILFP) framework, which integrates local and global features. A new deep learning architecture named modified XceptionNet (m-XceptionNet) is also proposed in this work, which extracts the global features effectively with lesser complexity. The proposed framework gives better performance in Rank1 metric for Market1501 (96.15%), CUHK03 (82.29%) and the newly created NEC01 (96.66%) datasets than the existing works. The mean Average Precision (mAP) calculated using the proposed framework gives 92%, 85% and 98%, respectively, for the same datasets.

• The Journal of the Convergence on Culture Technology
• /
• v.2 no.4
• /
• pp.71-76
• /
• 2016

In this study, de-identification policies of the US, the UK, Japan, China and Korea are compared to suggest a future direction of de-identification regulations and a method for vitalizing the big data industry. Efficiently using the de-identification technology and the standard of adequacy evaluation contributes to using personal information for the industry to develop services and technology while not violating the right of private lives and avoiding the restrictions specified in the Personal Information Protection Act. As a counteraction, the re-identification issue may occur, for re-identifying each person as a de-identified data collection. From the perspective of business, it is necessary to mitigate schemes for discarding some regulations and using big data, and also necessary to strengthen security and refine regulations from the perspective of information security.

• Journal of Digital Convergence
• /
• v.11 no.10
• /
• pp.737-746
• /
• 2013

This research aim is to investigate the rate and pattern of re-using the four-digit personal identification numbers(PINs). 1313 types of PINs were observed by 224 students who took this author's classes from last 2006 to 2011 at A-university. Some students used PINs as few as 3-4 and as many as 12-13. The average is 5.86 per person. The rates of re-using PINs were calculated by each student. 87%(195/224) of students reused PINs and 64% of them reused with just only one type of PINs, 20% reused with 2 types, and about 3% reused with 3-4 types. With respect to PINs, 884 out of the total 1313 PINs were reused, that is around 64.3%. In a broad sense, a pair of slight modification of PINs were also observed, that is, new PINs were partly matched in position or size of numbers of previous PINs. And if the reuse rate falling under the slight modification of PINs, 10.4% is added, about 75% of the PINs were reused in a broad sense. The re-using rate of male students is higher than the one of female students. This paper's results may provide to make plan for hacks of passwords.

• Proceedings of the KIEE Conference
• /
• 2003.11c
• /
• pp.532-535
• /
• 2003

In this paper, we propose a new algorithm that can enhance the speed of template matching of hand vascular pattern person verification or recognition system. Various template matching algorithms have advantages in the matching accuracy, but most of the algorithms suffer from computational burden. To reduce the computational amount, with accuracy maintained, we propose following template matching scenario as follows. firstly, original hand vascular image is re-sampled in order to reduce spatial resolution. Secondly, reconstructed image is projected to vertical and horizontal direction, being converted to two one dimensional (1D) data. Thirdly, converted data is used to estimate spatial discrepancy between stored template image and target image. Finally, matching begins from where the estimated order is highest, and finishes when matching decision function is computed to be over certain threshold. We've applied the proposed algorithm to hand vascular pattern identification application for biometrics, and observed dramatic matching speed enhancement. This paper presents detailed explanation of the proposed algorithm and evaluation results.

• Journal of the Korea Institute of Information Security & Cryptology
• /
• v.23 no.5
• /
• pp.977-983
• /
• 2013

This paper presents result of researching personal information exposure of Korean twitter and facebook users. Personally identifiable information such as e-mail and phone numer is exposed in the accounts less than 1%. However there are many cases that a person is identified by non personally identifiable information. For example, 350 thousands accounts are distinguished with other accounts because its name is unique. Using combination of information such as name and high school, we can distinguish 2.97 millions accounts. We also found 170 thousands account pairs that are candidate of one users' own account. Linkability between two accounts in two different domains means that the person is identified. Currently, only personally identifiable information is protected by policy. This paper shows that the policy has limited effects under the circumstances that a person can be identified by non personally identifiable information and the account linking is possible.