• 제목/요약/키워드: sequence-to-sequence model

검색결과 1,626건 처리시간 0.031초

Video Content Manipulation Using 3D Analysis for MPEG-4

  • Sull, Sanghoon
    • 방송공학회논문지
    • /
    • 제2권2호
    • /
    • pp.125-135
    • /
    • 1997
  • This paper is concerned with realistic mainpulation of content in video sequences. Manipulation of content in video sequences is one of the content-based functionalities for MPEG-4 Visual standard. We present an approach to synthesizing video sequences by using the intermediate outputs of three-dimensional (3D) motion and depth analysis. For concreteness, we focus on video showing 3D motion of an observer relative to a scene containing planar runways (or roads). We first present a simple runway (or road) model. Then, we describe a method of identifying the runway (or road) boundary in the image using the Point of Heading Direction (PHD) which is defined as the image of, the ray along which a camera moves. The 3D motion of the camera is obtained from one of the existing 3D analysis methods. Then, a video sequence containing a runway is manipulated by (i) coloring the scene part above a vanishing line, say blue, to show sky, (ii) filling in the occluded scene parts, and (iii) overlaying the identified runway edges and placing yellow disks in them, simulating lights. Experimental results for a real video sequence are presented.

  • PDF

On the Code Selection of a Multicode DS/CDMA System for a High Data Rate Transmission

  • Kim, Jin-Young
    • 대한전자공학회:학술대회논문집
    • /
    • 대한전자공학회 2000년도 추계종합학술대회 논문집(1)
    • /
    • pp.457-460
    • /
    • 2000
  • The effect of code selection for a multicode DS/CDMA system is evaluated for a high deta rate transmission, The performance is evaluated in terms of bit error and outage probabilities. The multipath fading channel is modeled as a Nakagami-m distribution which has been known to be appropriate to model the multipath fading in urban as well as indoor channels. From simulation results, it is shown that the concatenated sequence of Walsh code and Gold sequence is most promising among many code selections. The considerations in this paper can be applied to the next-generation mobile communication systems such as IMT-2000 which requires high bit rate transmissions.

  • PDF

다중 경로 채널 시스템에서 신경회로망을 이용한 간섭 신호 제거 (Rejection of Interference Signal Using Neural Network in Multi-path Channel Systems)

  • 석경휴
    • 한국음향학회:학술대회논문집
    • /
    • 한국음향학회 1998년도 학술발표대회 논문집 제17권 1호
    • /
    • pp.357-360
    • /
    • 1998
  • DS/CDMA system rejected narrow-band interference and additional White Gaussian noise which are occured at multipath, intentional jammer and multiuser to share same bandwidth in mobile communication systems. Because of having not sufficiently obtained processing gain which is related to system performance, they were not effectively suppressed. In this paper, an matched filter channel model using backpropagation neural network based on complex multilayer perceptron is presented for suppressing interference of narrow-band of direct sequence spread spectrum receiver in DS/CDMA mobile communication systems. Recursive least square backpropagation algorithm with backpropagation error is used for fast convergence and better performance in matched filter receiver scheme. According to signal noise ratio and transmission power ratio, computer simulation results show that bit error ratio of matched filter using backpropagation neural network improved than that of RAKE receiver of direct sequence spread spectrum considering of con-channel and narrow-band interference.

  • PDF

A Study on Identification of Nonminimum Phase Stable System from Partial Impulse Response Sequences

  • Lee, Won-Cheol;Bae, Myung-Jin;Im, Sung-Bin
    • The Journal of the Acoustical Society of Korea
    • /
    • 제15권1E호
    • /
    • pp.45-58
    • /
    • 1996
  • This paper addresses the problem of identifying the class of all stable system transfer functions that interpolate the given partial impulse response sequence. In this context, classical Pade approximations that are also stable, are shown to be a special case of this general formulation. The theory developed in this connection is utilized to obtain a new criterion for determining the model order and system parameters for rational systems, and, further, to generate nonminimum phase optimal stable rational approximatinos of nonrational systems from its impulse response sequence.

  • PDF

Construction of an RNase P Ribozyme Library System for Functional Genomics Applications

  • Hong, Sun-Woo;Choi, Hyo-Jei;Lee, Young-Hoon;Lee, Dong-Ki
    • Genomics & Informatics
    • /
    • 제5권1호
    • /
    • pp.6-9
    • /
    • 2007
  • An RNase P ribozyme library has been developed as a tool for functional genomics studies. Each clone of this library contains a random 18-mer and the sequence of M1 RNA, the catalytic subunit of RNase P. Repression of target gene expression is thus achieved by the complementary binding of mRNA to the random guide sequence and the successive target cleavage via M1 RNA. Cellular expression of the ribozyme expression was confirmed, and EGFP mRNA was used as a model to demonstrate that the RNase P ribozyme expression system can inhibit the target gene expression. The constructed RNase P ribozyme library has a complexity of $1.4\times10^7$. This novel library system should become a useful in functional genomics, to identify novel gene functions in mammalian cells.

Immunological Recognition by Artificial Neural Networks

  • Xu, Jin;Jo, Junghyo
    • Journal of the Korean Physical Society
    • /
    • 제73권12호
    • /
    • pp.1908-1917
    • /
    • 2018
  • The binding affinity between the T-cell receptors (TCRs) and antigenic peptides mainly determines immunological recognition. It is not a trivial task that T cells identify the digital sequences of peptide amino acids by simply relying on the integrated binding affinity between TCRs and antigenic peptides. To address this problem, we examine whether the affinity-based discrimination of peptide sequences is learnable and generalizable by artificial neural networks (ANNs) that process the digital experimental amino acid sequence information of receptors and peptides. A pair of TCR and peptide sequences correspond to the input for ANNs, while the success or failure of the immunological recognition correspond to the output. The output is obtained by both theoretical model and experimental data. In either case, we confirmed that ANNs could learn the immunological recognition. We also found that a homogenized encoding of amino acid sequence was more effective for the supervised learning task.

Mechanical behavior of outer square inner circular concrete-filled dual steel tubular stub columns

  • Ding, Fa-xing;Wang, Wenjun;Liu, Xue-mei;Wang, Liping;Sun, Yi
    • Steel and Composite Structures
    • /
    • 제38권3호
    • /
    • pp.305-317
    • /
    • 2021
  • The mechanical behavior of the outer square inner circular concrete-filled dual steel tubular (SCCFT) stub columns under axial compression is investigated by means of experimental research, numerical analysis and theoretical investigation. Parameters such as diameter ratio, concrete strength and steel ratio were discussed to identify their influence on the mechanical properties of SCCFT short columns on the basis of the experimental investigation of seven SCCFT short columns. By establishing a finite element model, nonlinear analysis was performed to discuss the longitudinal and transverse stress of the dual steel tubes. The longitudinal stress characteristics of the core and sandwich concrete were also analyzed. Furthermore, the failure sequence was illustrated and the reasonable cross-section composition of SCCFT stub column was proposed. A formula to predict the axial load capacity of SCCFT stub column was advanced and verified by the results from experiment and the finite element.

단방향 및 양방향 순환 신경망의 성능 평가 (Performance Evaluation of Unidirectional and Bidirectional Recurrent Neural Networks)

  • ;정경희 ;추현승
    • 한국정보처리학회:학술대회논문집
    • /
    • 한국정보처리학회 2023년도 춘계학술발표대회
    • /
    • pp.652-654
    • /
    • 2023
  • The accurate prediction of User Equipment (UE) paths in wireless networks is crucial for improving handover mechanisms and optimizing network performance, particularly in the context of Beyond 5G and 6G networks. This paper presents a comprehensive evaluation of unidirectional and bidirectional recurrent neural network (RNN) architectures for UE path prediction. The study employs a sequence-to-sequence model designed to forecast user paths in a wireless network environment, comparing the performance of unidirectional and bidirectional RNNs. Through extensive experimentation, the paper highlights the strengths and weaknesses of each RNN architecture in terms of prediction accuracy and computational efficiency. These insights contribute to the development of more effective predictive path-based mobility management strategies, capable of addressing the challenges posed by ultra-dense cell deployments and complex network dynamics.

Control and aggregation (II)

  • Han, Sung-Shin
    • Journal of the Korean Statistical Society
    • /
    • 제9권1호
    • /
    • pp.39-60
    • /
    • 1980
  • In the last paper, we have discussed the cannonical representatin of a dynamic linear model, on which some aggregation schemes were devised. The relationships of those aggregation schemes with dynamic properties were investigated. This paper tries to analyse she control strategy for the aggregated linear dynamic model and to investigate the dynamic properties of disaggregative model controlled by aggregated model. For the logical consistency with the last paper, all the sections and all the equations are numbered in a sequence.

  • PDF

목적 지향 대화를 위한 효율적 질의 의도 분석에 관한 연구 (Effective Text Question Analysis for Goal-oriented Dialogue)

  • 김학동;고명현;임헌영;이유림;지민규;김원일
    • 방송공학회논문지
    • /
    • 제24권1호
    • /
    • pp.48-57
    • /
    • 2019
  • 본 연구는 목적 지향 대화 시스템 내에서 단일 한국어 텍스트 형식의 질문으로부터 질의자의 의도를 파악하는 것을 목적으로 한다. 목적 지향 대화 시스템은 텍스트 또는 음성을 통한 사용자의 특수한 요구를 만족시켜주는 대화 시스템을 의미한다. 의도 분석 과정은 답변 생성에 앞서 사용자의 질의 의도를 파악하는 단계로, 목적 지향 대화 시스템 전체의 성능에 큰 영향을 준다. 생활화학제품이라는 특정 분야에 제안 모델을 사용하였고, 해당 분야와 관련된 한국어 텍스트 데이터를 이용하였다. 특정 분야에 독립적이며 범용적인 의도를 의미하는 화행과, 특정 분야에 종속적인 의도를 의미하는 개념열로 나누어 분석한다. 화행과 개념열을 분석하기 위하여 단어 임베딩 모델, 합성곱 신경망을 이용한 분류 방법을 제안한다. 단어 임베딩 모델을 통하여 단어의 의미정보를 추상화하고, 추상화된 단어의 의미정보를 기반으로 합성곱 신경망을 통하여 개념열 및 화행 분류를 수행한다.