동영상의 유해성 판별을 위한 대표 프레임 선정 기법

김성균,박명철,하석운,Kim, Seong-Gyun,Park, Myeong-Chul,Ha, Seok-Wun 한국정보통신학회 2006 한국정보통신학회논문지 Vol.10 No.10

동영상에서 유해정보를 판별하기 위해서는 대표프레임을 효율적으로 선정 할 수 있어야 한다. 기존의 대표프레임 선정 기법은 대부분이 장면전환을 중심으로 이루어진다. 이러한 기법 은 연속된 변화특성을 가지는 유해 동영상의 경우에는 불필요한 대표프레임으로 인해 전체적인 판별효율을 저하시킨다. 본 논문에서는 판별시스템의 입력이 되는 대표프레임을 프레 임간 변화특성을 이용하여 선정하는 기법을 제안한다. 이 기법의 실험을 위해서 기존의 판별시스템에 제안된 기법으로 선정된 대표프레임을 투입한 경우에 90% 이상이 유해하다고 판별하여 입력의 적합성이 입증되었으며, 선정된 대표프레임의 수도 I-프레임 비 해 68%의 감소율을 보여 시간적 효율성도 입증되었다. 그러므로 본 기법은 효율적인 유해성 판별시스템을 가능하게 하여, 건전한 동영상 정보의 유통에 효과적으로 기여할 수 있다. A key-frame should be efficiently selected to distinguish bad information from the videos. A previous technique selecting a key-frame mostly consists of the transformation scene-centered. In the case of harmful videos containing the quaility of continuous changes, the technique makes the total rate be reduced by an unnecessary key-frame. This thesis suggests the technique selecting a key-frame, an entry of the distinguishing system by using the quality of changes between the frames. In the experiment of this technique, it was proved that over 90% of the bad information was distinguished by the selected key frame, and also time efficiency was proved by showing 68% of decrement compared to the numbers I-frame. Therefore, This technique makes the system efficient to distinguish bad information, and efficiently can contribute to the distribution of the healthy movie information.

Why Gabor Frames? Two Fundamental Measures of Coherence and Their Role in Model Selection

Bajwa, Waheed U.,Calderbank, Robert,Jafarpour, Sina The Korea Institute of Information and Commucation 2010 Journal of communications and networks Vol.12 No.4

The problem of model selection arises in a number of contexts, such as subset selection in linear regression, estimation of structures in graphical models, and signal denoising. This paper studies non-asymptotic model selection for the general case of arbitrary (random or deterministic) design matrices and arbitrary nonzero entries of the signal. In this regard, it generalizes the notion of incoherence in the existing literature on model selection and introduces two fundamental measures of coherence-termed as the worst-case coherence and the average coherence-among the columns of a design matrix. It utilizes these two measures of coherence to provide an in-depth analysis of a simple, model-order agnostic one-step thresholding (OST) algorithm for model selection and proves that OST is feasible for exact as well as partial model selection as long as the design matrix obeys an easily verifiable property, which is termed as the coherence property. One of the key insights offered by the ensuing analysis in this regard is that OST can successfully carry out model selection even when methods based on convex optimization such as the lasso fail due to the rank deficiency of the submatrices of the design matrix. In addition, the paper establishes that if the design matrix has reasonably small worst-case and average coherence then OST performs near-optimally when either (i) the energy of any nonzero entry of the signal is close to the average signal energy per nonzero entry or (ii) the signal-to-noise ratio in the measurement system is not too high. Finally, two other key contributions of the paper are that (i) it provides bounds on the average coherence of Gaussian matrices and Gabor frames, and (ii) it extends the results on model selection using OST to low-complexity, model-order agnostic recovery of sparse signals with arbitrary nonzero entries. In particular, this part of the analysis in the paper implies that an Alltop Gabor frame together with OST can successfully carry out model selection and recovery of sparse signals irrespective of the phases of the nonzero entries even if the number of nonzero entries scales almost linearly with the number of rows of the Alltop Gabor frame.

Why Gabor Frames? Two Fundamental Measures of Coherence and Their Role in Model Selection (Invited Paper)

Waheed U. Bajwa,Robert Calderbank,Sina Jafarpour 한국통신학회 2010 Journal of communications and networks Vol.12 No.4

The problem of model selection arises in a number of contexts, such as subset selection in linear regression, estimation of structures in graphical models, and signal denoising. This paper studies non-asymptotic model selection for the general case of arbitrary (random or deterministic) design matrices and arbitrary nonzero entries of the signal. In this regard, it generalizes the notion of incoherence in the existing literature on model selection and introduces two fundamental measures of coherence—termed as the worst-case coherence and the average coherence—among the columns of a design matrix. It utilizes these two measures of coherence to provide an in-depth analysis of a simple, model-order agnostic one-step thresholding (OST) algorithm for model selection and proves that OST is feasible for exact as well as partial model selection as long as the design matrix obeys an easily verifiable property, which is termed as the coherence property. One of the key insights offered by the ensuing analysis in this regard is that OST can successfully carry out model selection even when methods based on convex optimization such as the lasso fail due to the rank deficiency of the submatrices of the design matrix. In addition, the paper establishes that if the design matrix has reasonably small worst-case and average coherence then OST performs near-optimally when either (i) the energy of any nonzero entry of the signal is close to the average signal energy per nonzero entry or (ii) the signal-to-noise ratio in the measurement system is not too high. Finally, two other key contributions of the paper are that (i)it provides bounds on the average coherence of Gaussian matrices and Gabor frames, and (ii) it extends the results on model selection using OST to low-complexity, model-order agnostic recovery of sparse signals with arbitrary nonzero entries. In particular, this part of the analysis in the paper implies that an AlltopGabor frame together with OST can successfully carry out model selection and recovery of sparse signals irrespective of the phases of the nonzero entries even if the number of nonzero entries scales almost linearly with the number of rows of the Alltop Gabor frame.

H.264/AVC에서의 효율적인 고속 다중 참조 프레임 선택 기법

이현우(Hyunwoo Lee),유종민(Jongmin Ryu),정제창(Jechang Jeong) 한국통신학회 2008 韓國通信學會論文誌 Vol.33 No.10C

H.264/AVC 동영상 부호화 표준은 높은 부호화 효율을 위해서 가변 블록 크기 부호화, 1/4 화소 정확도의 움직임 예측, 다중 참조 프레임 사용, 율-왜곡 최적화 등과 같은 기법을 도입하였다. 그러나 이러한 부호화 방법들은 움직임 예측에 필요한 복잡도를 크게 증가시키는 단점이 있다. 특히 다중 참조 프레임 움직임 예측을 수행함으로써 증가되는 계산량은 탐색되는 참조 프레임의 개수에 따라 선형적으로 비례하여 늘어난다. 따라서 본 논문에서는 다중 참조 프레임 움직임 예측에서 탐색되는 참조 프레임의 개수를 효율적으로 조절함으로써 복잡도를 크게 줄일 수 있는 방법을 제안한다. 제안하는 알고리듬은 인터 16×16 모드와 인접 블록의 최적 참조 프레임 정보를 이용하여 나머지 화면간 모드들에서의 불필요한 탐색 과정을 생략한다. 실험 결과는 제안하는 방법을 적용함으로써 영상의 화질과 비트율의 큰 변화 없이 부호화 수행 시간을 평균 58.54% 정도의 큰 폭으로 감소시킴을 보인다. 또한 이 방법은 다른 고속 움직임 예측 알고리듬을 같이 적용함으로써 추가적인 성능 향상을 얻을 수 있는 확장성을 가진다. In order to achieve high coding efficiency, H.264/AVC video coding standard adopts the techniques such as variable block size coding, motion estimation with quarter-pel precision, multiple reference frames, rate-distortion optimization, and etc. However, these coding methods have a defect to greatly increase the complexity for motion estimation. Particularly, from multiple reference frame motion estimation, the computational burden increases in proportion to the number of the searched reference frames. Therefore, we propose the method to reduce the complexity by controlling the number of the searched reference frames in motion estimation. Proposed algorithm uses the optimal reference frame information in both P16×16 mode and the adjacent blocks, thus omits unnecessary searching process in the rest of inter modes. Experimental results show the proposed method can save an average of 57.31% of the coding time with negligible quality and bit-rate difference. This method also can be adopted with any of the existing motion estimation algorithm. Therefore, additional performance improvement can be obtained.

A Novel Selective Frame Discard Method for 3D Video over IP Networks

( Young-uk Chung ) 한국인터넷정보학회 2010 KSII Transactions on Internet and Information Syst Vol.4 No.6

Three dimensional (3D) video is expected to be an important application for broadcast and IP streaming services. One of the main limitations for the transmission of 3D video over IP networks is network bandwidth mismatch due to the large size of 3D data, which causes fatal decoding errors and mosaic-like damage. This paper presents a novel selective frame discard method to address the problem. The main idea of the proposed method is the symmetrical discard of the two dimensional (2D) video frame and the depth map frame. Also, the frames to be discarded are selected after additional consideration of the playback deadline, the network bandwidth, and the inter-frame dependency relationship within a group of pictures (GOP). It enables the efficient utilization of the network bandwidth and high quality 3D IPTV service. The simulation results demonstrate that the proposed method enhances the media quality of 3D video streaming even in the case of bad network conditions.

Optimum Sizing Design of Steel Frame Structures Using Evaporation Rate-Based Water Cycle Algorithm with Greedy Selection

Serdar Carbas 한국강구조학회 2022 International Journal of Steel Structures Vol.22 No.4

The evaporation rate-based water cycle algorithm (WCA-ER) is one of the famous nature-inspired metaheuristic optimization methods asserted to simulate principal idea behind observation of water cycle process with evaporation rate where how rivers and streams fl ow to the sea. In WCA-ER, an evaporation rate notion was identifi ed formatively to standard WCA for achieving a better balance between the exploration and exploitation phases. This enables making better choice decision. Furthermore, in this study the WCA-ER is equipped with a greedy selection scheme bringing about much effi cient search technique for obtaining the optimal designs by maintaining solution diversity and enhancing the exploitation phase. So, the main objective of this paper is to attain optimal sizing designs of both planar and space steel frame structures using WCA-ER with greedy selection. In this context, the WCA-ER with greedy selection choses the W steel sections to assign the member groups treated as design variables. Not only the strength and stability constraints imposing from LRFD-AISC code of practice design, but also the defl ection, drift, and geometric constraints are taken into account in design optimization. The optimally frame designs obtained by WCA-ER with greedy selection are tabulated and they are compared with previously reported optimum solutions via other metaheuristic algorithms. The achieved best feasible steel frame structures for design examples illustrate the robustness and effi ciency of the algorithmic performance of the WCA-ER with greedy selection.

한국어 유의어 변별 교육 : 틀 의미론 적용을 중심으로

마릉연(Ma ling yan) 우리말교육현장학회 2015 우리말교육현장연구 Vol.9 No.1

This article analyzes synonymy semantics according the theories frame. Traditional semantics analyzes language internally and ignores the human experiences and cognition mechanisms. Cognitive Semantics argues that the meaning of words lies not in language itself but in the frames constructed by a series of cognition activities. The establishment of meaning should take encyclopedic knowledge into account. Frame Semantics provides a new perspective to the explanation of meaning, and it can be used to distinguish synonymies from three perspectives of "profiling", 'selectional restruction', 'Linking of the framework elements and gyeokteul' And using the three methods, the synonyms for "넘다-건너다" were analyzed. According to the analysis result, proposed the education plan for Chinese advanced learners. 전통의미론은 어휘의 의미 구조에 대한 분석을 통해 유의어에 접근하는데, 이러한 분석 방법은 백과사전적 지식(encyclopedic knowledge)이 결여된다는 단점이 있다. 이에 본고는 틀 의미론을 도입하여 유의어의 의미 해석과 변별에 틀 의미론을 활용할 필요성을 살펴보고, 유의어의 변별에 '윤곽부여', '틀 선택제약', '틀 요소와 격틀 링킹'의 3가지 방법을 제시하였다. 그리고 이 3가지 방법을 이동 동사 유의어 '건너다', '넘다'에 적용하여 분석하였다. 다음 이 분석 결과를 바탕으로 하여 중국인 한국어 학습자를 대상으로 한 학습자 중심 교수 방안을 제안하였다. 이러한 교수 · 학습 방법은 학습자가 주도적으로 개념체계를 형성해 나감으로써 궁극적으로 자신의 의사소통 능력을 향상시키는 데 유용할 것으로 기대한다.

움직임 벡터 참조 지도 기반의 고속 참조 영상 선택 방법

이경희(Kyung-Hee Lee),고만근(Man-Geun Ko),서보석(Bo-Seok Seo),서재원(Jae-Won Suh) 한국콘텐츠학회 2010 한국콘텐츠학회논문지 Vol.10 No.4

다중 참조 영상을 이용한 가변 블록 크기의 움직임 예측 및 보상 기법이 부호화 효율을 높이기 위해 H.264/AVC에 채택되었다. 하지만 움직임 예측 및 보상으로 인한 계산량은 다중 참조 영상과 가변 블록의 수에 비례하여 증가한다. 본 논문에서는 화질은 유지하며 계산량을 줄이기 위한 새로운 고속 참조 영상선택 방법을 제안한다. 먼저 다중 참조 영상들에 대해서 4×4 블록의 SAD값을 이용하여 움직임 벡터 참조지도를 만든다. 다음으로 가변 블록 크기의 움직임 예측 및 보상이 움직임 벡터 참조 지도를 활용하여 실시된다. 제안하는 방법은 H.264/AVC 표준과 비교하여 BDPSNR은 평균적으로 0.01dB 나빠지고 BDBR은 약 0.27% 증가하지만 영상 부호화 속도를 약 38% 단축시킨다. The variable block size motion estimation (ME) and compensation (MC) using multiple reference frames is adopted in H.264/AVC to improve coding efficiency. However, the computational complexity for ME/MC increases proportional to the number of reference frames and variable blocks. In this paper, we propose a new efficient reference frame selection algorithm to reduce the complexity while keeping the visual quality. First, a motion vector reference map is constructed by SAD of 4×4 block unit for multi reference frames. Next, the variable block size motion estimation and motion compensation is performed according to the motion vector reference map. The computer simulation results show that the average loss of BDPSNR is -0.01dB, the increment of BDBR is 0.27%, and the encoding time is reduced by 38% compared with the original method for H.264/AVC.

동영상 부호화의 움직임 추정 및 보상을 위한 고속 다중 참조 프레임 선택 기법

김재훈(Jae-Hoon Kim),김명진(Myoung-Jin Kim),홍민철(Min-Cheol Hong) 한국통신학회 2007 韓國通信學會論文誌 Vol.32 No.11C

본 논문에서는 동영상 부호화 방식의 움직임 추정을 위한 고속 다중 참조 프레임 선택 기법에 대해 제안한다. 움직임 추정을 위해 다중 프레임을 참조하는 H.264 동영상 부호화 방식의 경우, 블록별로 선택된 최적의 참조 프레임들은 일정한 상관관계를 유지하는 특성이 존재한다. 제안 방식은 이러한 통계적 특성을 이용하여, 임의의 블록 크기의 움직임 추정 모드에서 선택된 최적의 참조 프레임 정보를 이용하여 하위 블록 크기의 움직임 추정 모드에서 사용할 참조 프레임을 예측한다. 실험 결과를 통해 5개의 참조 프레임 사용을 기준으로 제안하는 기법을 사용하였을 때, 움직임 추정에 소요되는 시간은 가용한 모든 참조 프레임에 대해 움직임 추정을 수행하는 방식 대비 평균 60% 감소되었으며, PSNR 및 발생 비트율 측면에서 거의 동일한 성능을 유지함을 확인할 수 있었다. In this paper, we propose a fast multiple reference frame selection method for motion estimation and compensation in video coding. Reference frames selected as an optimal reference frame by variable block sizes motion estimation have the statistical characteristic that was based on block size. Using the statistical characteristic, reference frames for smaller block size motion estimation can be selected from reference frame which was decided as an optimal one for the upper layer block size. Simulation results show that the proposal method decreased the computations about 60%. Nevertheless, PSNR and bit rate were almost same as the performances of original H.264 multiple reference motion estimation.

비디오 프레임 선택을 통한 주거 공간 인간 행동 인식 모델 경량화 방안 제안

김소현,한지형 한국정보과학회 2023 정보과학회논문지 Vol.50 No.12

Residential area closed-circuit televisions (CCTVs) need human action recognition (HAR) to predict any accidents and crucial problems. HAR model must be not only accurate but also light and fast to apply in the real world. Therefore, in this paper, a cross-modal PoseC3D model with a frame selection method is proposed. The proposed cross-modal PoseC3D model integrates multi-modality inputs (i.e., RGB image and human skeleton data) and trains them in a single model. Thus, the proposed model is lighter and faster than previous works such as two-pathway PoseC3D. Moreover, we apply the frame selection method to use only the meaningful frames based on differences between frames instead of using the whole frame of a video. AI Hub open dataset was used to verify the performance of proposed method. The experimental results showed that the proposed method achieves similar or better performance and is much lighter and faster than those in the previous works.