• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.13 no.12
• /
• pp.6159-6174
• /
• 2019

Recently, Kim et al. introduced the Hamming + k with m overlapped pixels data hiding (Hk_mDH) based on matrix encoding. The embedding rate (ER) of this method is 0.54, which is better than Hamming code HC (n, n - k) and HC (n, n - k) +1 DH (H1DH), but not enough. Hamming code data hiding (HDH) is using a covering function COV(1, n = 2^k -1, k) and H1DH has a better embedding efficiency, when compared with HDH. The demerit of this method is that they do not exploit their space of pixels enough to increase ER. In this paper, we increase ER using sequential Hk_mDH (SHk_mDH ) through fully exploiting every pixel in a cover image. In SHk_mDH, a collision maybe happens when the position of two pixels within overlapped two blocks is the same. To solve the collision problem, in this paper, we have devised that the number of modification does not exceed 2 bits even if a collision occurs by using OPAP and LSB. Theoretical estimations of the average mean square error (AMSE) for these schemes demonstrate the advantage of our SHk_mDH scheme. Experimental results show that the proposed method is superior to previous schemes.

• Proceedings of the IEEK Conference
• /
• 2002.07a
• /
• pp.152-155
• /
• 2002

In this paper, a new approach is proposed for the error detection dedicated to MPEG-4 video coding by using the digital watermarking. In the process of encoding, the proposed scheme abstracts the macroblock features of headers, motion vectors, and Discrete Cosine Transform (DCT) coefficients, which are embedded in the quantized DCT coefficients as the digital watermarks. The decoder performs the accurate error detection through the watermark extraction. Simulation results demonstrate that the error detection ability of the proposed approach can be significantly enhanced, and that digital watermark embedding incurs little degradation to the picture quality.

• Proceedings of the Korean Society for Noise and Vibration Engineering Conference
• /
• 2002.11b
• /
• pp.301-305
• /
• 2002

Among many structural dynamics modification methods for plate and shell vibration problems, embedding an emboss to the surface is very efficient. But deciding an optimal position and shape using optimization algorithm needs defining geometry and remeshing the model for every iteration step to implement the method, which takes much numerical cost. An equivalent element produced here lessen the cost by representing the geometrical characteristics of an emboss using the element's material properties and thickness becoming a geometrically homogenous element of the base plate or shell. Some efficient factors which let the equivalent system have the same dynamical response as the original system embedded with emboss will be shown and the degree of equivalence will be tested in terms of natural frequency matching.

• Proceedings of the KIEE Conference
• /
• 2004.11b
• /
• pp.91-93
• /
• 2004

Load forecasting is essential in order to supply electrical energy stably and economically in power systems. ANNs have flexibility to predict a nonlinear feature of load profiles. In this paper, we selected just the necessary input variables used in the paper(2) which is based on the phase-space embedding of a load time-series and reviewing others. So only 5 input variables were selected to forecast for spring, fall and winter season and another input considering temperature sensitivity is added during the summer season. The training cases are also selected from all previous data composed training cases of a 7-day, 14-day and 30-day period. Finally, we selected the training case of a 7-day period because it can be used in STLF without sacrificing the accuracy of the forecast. This allows more compact ANNs, smaller training cases. Consequently, test results show that compact neural networks can be forecasted without sacrificing the accuracy.

• Journal of Korea Multimedia Society
• /
• v.11 no.6
• /
• pp.796-806
• /
• 2008

This paper presents a blind video watermarking algorithm that has the robustness against spatial, temporal, and SNR scalability and transcoding for the copyright protection of video contents in heterogeneous multimedia service. The proposed process of watermark embedding and detecting is accomplished on base layer for considering spatial scalability. The watermark consists of the string and the ordering number of string for considering temporal scalability. Thus, each of frames has the bitstream of one character and a ordering number of its character. To robust against FGS, the proposed algorithm quantizes low and middle frequency coefficients in ROI region of each of frames and embeds its watermark bitstream into the specific bits of the quantized coefficients. Experimental results verified that the proposed algorithm satisfies the invisibility of watermark and also has the robustness against spatial scalability, temporal scalability and FGS.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2018.06a
• /
• pp.329-331
• /
• 2018

지속적인 딥러닝 기반의 영상처리 기술의 발전으로 객체분류나 객체검출 문제에 대해서 뛰어난 성능 보이고 있다. 하지만 객체추적 문제에서는 성능이 좋은 추적기는 실시간 동작이 불가능하고 딥러닝 기반의 객체추적도 단일 객체에만 고려한 기법이 많기 때문에 개선할 필요가 있다. 전처리로 검출된 객체영역과 kalman filter를 통해 예측된 추적영역 간의 embedding feature 비교를 통해 동일인물인지 판단하여 고유 ID를 부여하고 추적한다. 객체끼리 교차하거나 가려지는 상황에서 추적을 실패하게 되는데 이 후에 지속적인 추적을 위해 IoU 비교를 통해 후보 추적기로 남겨두는 과정을 거친다. 실험 결과 실시간 동작여부와 객체끼리 교차하거나 프레임 밖으로 나갔다가 다시 나타나는 경우에도 추적이 가능함을 확인하였다.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2022.11a
• /
• pp.106-108
• /
• 2022

이미지 기반 3D 형상 복원에 있어서, 이미지에 보이지 않는 폐색(Occlusion) 영역 부분에 대한 정보가 손실되므로 완전한 복원에 어려움이 있으며, 세밀한 텍스쳐(Texture) 표현이 이루어지지 않고 심한 평활화(Smoothing)나 고립된 노이즈 메쉬(Isolated Noise Mesh) 등 구조적 훼손이 발생한다. 주로 깊은 신경망을 이용하여, 음함수(Implicit Function) 기반 방법은 사전훈련이 완료된 보조 신경망들을 전면부에 배치하거나, Hourglass 등 임베딩(Embedding) 아키텍처를 추가하거나, 또는 표면 법선(Surface Normal)과 같은 환시(Hallucination)를 생성하여 신경망에 입력하기도 한다. 이 논문에서는, 인물의 이미지를 입력받아 색상, 머리카락 및 의상을 포함하는 완전 3D 인간 복원 기술들을 조망해본다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2021.05a
• /
• pp.290-292
• /
• 2021

최근 인코더-디코더 구조의 자연어 처리모델이 활발하게 연구가 이루어지고 있다. 인코더-디코더기반의 언어모델은 특히 본문의 내용을 새로운 문장으로 요약하는 추상(Abstractive) 요약 분야에서 널리 사용된다. 그러나 기존의 언어모델은 단일 문서 및 문장을 전제로 설계되었기 때문에 기존의 언어모델에 다중 문장을 요약을 적용하기 어렵고 주제가 다양한 여러 문장을 요약하면 요약의 성능이 떨어지는 문제가 있다. 따라서 본 논문에서는 다중 문장으로 대표적이고 상품 리뷰를 워드 임베딩의 유사도를 기준으로 클러스터를 구성하여 관련성이 높은 문장 별로 인공 신경망 기반 언어모델을 통해 요약을 수행한다. 제안하는 모델의 성능을 평가하기 위해 전체 문장과 요약 문장의 유사도를 측정하여 요약문이 원문의 정보를 얼마나 포함하는지 실험한다. 실험 결과 기존의 RNN 기반의 요약 모델보다 뛰어난 성능의 요약을 수행했다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2024.05a
• /
• pp.331-332
• /
• 2024

최근 디지털 워터마킹 기술은 디지털 콘텐츠의 저작권 보호 및 추적을 위해 활용되고 있다. 그러나 종래의 워터마킹 기술은 이미지에 워터마크 이미지 전체를 삽입하기 때문에 왜곡 공격에 취약하다. 이러한 문제를 해결하기 위해 본 연구에서는 워터마크 분할 삽입 기법을 제안하였다. 워터마크 분할 삽입 기법을 사용하면 종래 방법 대비 20%p의 손실률이 증가하더라도 원본 워터마크를 복구할 수 있어 1.5배 향상된 성능을 보인다.

• Journal of Korean Medicine Rehabilitation
• /
• v.33 no.3
• /
• pp.149-160
• /
• 2023

Objectives This study is for reporting current status and strategies of thread embedding acupuncture (TEA) in temporomandibular joint disorder (TMD) using web-based survey. Methods Survey was conducted online via E-mail among Korean medicine doctors registered with the Association of Korean Medicine. The questionnaire is developed by Korean medicine doctor in Kyung Hee University Hospital at Gangdong. The survey consisted mainly of multiple-choice questions on the current status and strategies of TEA for TMD. Results Total of 427 doctors responded. TEA was mostly used for 'Cosmetic purpose and others' with 287 respondents (67.2%), and 102 respondents (23.9%) having experience with TEA for TMD. The most common purpose for TEA for TMD was 'Improving muscle contraction and tension' with 290 respondents (67.9%). The average interval was reported 2.12 weeks, and the most common response for the number of treatments was five sessions with 127 respondents (29.7%). The most common criterion for determining the treatment site was based on anatomical structure, accounting for 92.7%. The most effective anatomical structure was the 'Masseter muscle' with 83.1%, followed by the 'Temporal muscle' with 51.8%. TEA direction for TMD was dominant in 'affected side' for all muscles. Conclusions Through a survey, we can investigate clinical usage of TEA for TMD. This study can be helpful in creating standardized criteria for TEA on TMD in the future.