• 제목/요약/키워드: Double Sequences

검색결과 136건 처리시간 0.023초

음성 인식을 위한 sequence-to-sequence 심층 신경망의 이중 attention 기법 (Double-attention mechanism of sequence-to-sequence deep neural networks for automatic speech recognition)

  • 육동석;임단;유인철
    • 한국음향학회지
    • /
    • 제39권5호
    • /
    • pp.476-482
    • /
    • 2020
  • 입력열과 출력열의 길이가 다른 경우 attention 기법을 이용한 sequence-to-sequence 심층 신경망이 우수한 성능을 보인다. 그러나, 출력열의 길이에 비해서 입력열의 길이가 너무 긴 경우, 그리고 하나의 출력값에 해당하는 입력열의 특성이 변화하는 경우, 하나의 문맥 벡터(context vector)를 사용하는 기존의 attention 방법은 적당하지 않을 수 있다. 본 논문에서는 이러한 문제를 해결하기 위해서 입력열의 왼쪽 부분과 오른쪽 부분을 각각 개별적으로 처리할 수 있는 두 개의 문맥 벡터를 사용하는 이중 attention 기법을 제안한다. 제안한 방법의 효율성은 TIMIT 데이터를 사용한 음성 인식 실험을 통하여 검증하였다.

MEAN CONVERGENCE THEOREMS AND WEAK LAWS OF LARGE NUMBERS FOR DOUBLE ARRAYS OF RANDOM ELEMENTS IN BANACH SPACES

  • Dung, Le Van;Tien, Nguyen Duy
    • 대한수학회보
    • /
    • 제47권3호
    • /
    • pp.467-482
    • /
    • 2010
  • For a double array of random elements {$V_{mn};m{\geq}1,\;n{\geq}1$} in a real separable Banach space, some mean convergence theorems and weak laws of large numbers are established. For the mean convergence results, conditions are provided under which $k_{mn}^{-\frac{1}{r}}\sum{{u_m}\atop{i=1}}\sum{{u_n}\atop{i=1}}(V_{ij}-E(V_{ij}|F_{ij})){\rightarrow}0$ in $L_r$ (0 < r < 2). The weak law results provide conditions for $k_{mn}^{-\frac{1}{r}}\sum{{T_m}\atop{i=1}}\sum{{\tau}_n\atop{j=1}}(V_{ij}-E(V_{ij}|F_{ij})){\rightarrow}0$ in probability where {$T_m;m\;{\geq}1$} and {${\tau}_n;n\;{\geq}1$} are sequences of positive integer-valued random variables, {$k_{mn};m{{\geq}}1,\;n{\geq}1$} is an array of positive integers. The sharpness of the results is illustrated by examples.

Vehicle trajectory prediction based on Hidden Markov Model

  • Ye, Ning;Zhang, Yingya;Wang, Ruchuan;Malekian, Reza
    • KSII Transactions on Internet and Information Systems (TIIS)
    • /
    • 제10권7호
    • /
    • pp.3150-3170
    • /
    • 2016
  • In Intelligent Transportation Systems (ITS), logistics distribution and mobile e-commerce, the real-time, accurate and reliable vehicle trajectory prediction has significant application value. Vehicle trajectory prediction can not only provide accurate location-based services, but also can monitor and predict traffic situation in advance, and then further recommend the optimal route for users. In this paper, firstly, we mine the double layers of hidden states of vehicle historical trajectories, and then determine the parameters of HMM (hidden Markov model) by historical data. Secondly, we adopt Viterbi algorithm to seek the double layers hidden states sequences corresponding to the just driven trajectory. Finally, we propose a new algorithm (DHMTP) for vehicle trajectory prediction based on the hidden Markov model of double layers hidden states, and predict the nearest neighbor unit of location information of the next k stages. The experimental results demonstrate that the prediction accuracy of the proposed algorithm is increased by 18.3% compared with TPMO algorithm and increased by 23.1% compared with Naive algorithm in aspect of predicting the next k phases' trajectories, especially when traffic flow is greater, such as this time from weekday morning to evening. Moreover, the time performance of DHMTP algorithm is also clearly improved compared with TPMO algorithm.

Cloning and Purification of Envelope Proteins (VP19, VP28) and Nucleocapsid Proteins (VP15, VP35) Genes of a Shrimp White Spot Syndrome Virus Isolates in Korea

  • Seok, Seung-hyeok;Park, Jae-hak
    • 한국수의병리학회:학술대회논문집
    • /
    • 한국수의병리학회 2003년도 추계학술대회초록집
    • /
    • pp.41-41
    • /
    • 2003
  • White spot syndrome virus (WSSV) is the causative agent of a disease that has led to severe mortalities of cultured shrimps in Korea and many other countries. Since 1993, massive mortalities due to the viral infection have also occurred in the penaeid shrimps cultured in Korea. WSSV is a large, circular, double stranded (ds) DNA virus and an enveloped, ellipsoid virus with a rod-shaped nucleocapsid with flat ends. In order to identify the characteristics of this Korean isolate of WSSV, the genes for four virion proteins, VP15, VP19, VP28 and VP35 were cloned and their sequences were compared with the available pool of WSSV gene sequences in the GenBank/EMBL databases. From these comparisons, we confirm the occurrence of WSSV in Korea and deduce that, VP15, VP28 and VP35 genes are identically conserved among the Korean isolate and geographically different foreign isolates, but VP19 amino acid sequences of the Korean WSSV isolates changed valine of the foreign isolates into aspartate. (omitted)

  • PDF

New Generalized PWM Schemes for Multilevel Inverters Providing Zero Common-Mode Voltage and Low Current Distortion

  • Nguyen, Nho-Van;Nguyen, Tam-Khanh Tu
    • Journal of Power Electronics
    • /
    • 제19권4호
    • /
    • pp.907-921
    • /
    • 2019
  • This paper presents two advanced hybrid pulse-width modulation (PWM) strategies for multilevel inverters (MLIs) that provide both common-mode voltage (CMV) elimination and current ripple reduction. The first PWM utilizes sequences that apply one switching state at the double ends of a half-carrier cycle. The second PWM combines the advantages of the former and an existing four-state PWM. Analyses of the harmonic characteristics of the two groups of switching sequences based on a general switching voltage model are carried out, and algorithms to optimize the current ripple are proposed. These methods are simple and can be implemented online for general n-level inverters. Using a three-level NPC inverter and a five-level CHB inverter, good performances in terms of the root mean square current ripple are obtained with the proposed PWM schemes as indicated through improved harmonic distortion factors when compared to existing schemes in almost the entire region of the modulation index. This also leads to a significant reduction in the current total harmonic distortion. Simulation and experimental results are provided to verify the effectiveness of the proposed PWM methods.

완두콩(Pisum sativum)에서 Ribulose-1,5-Bisphosphate Carboxylase Small Subunit 유전자의 cDNA 클로닝과 광유도성 발현에 관한 연구 (Cloning of cDNA Encoding the Precursor to the Small Subunit of Ribulose-1,5-Bisphosphate Carboxylase in Pea 9Pisum sativum))

  • 김한집
    • Journal of Plant Biology
    • /
    • 제32권1호
    • /
    • pp.33-40
    • /
    • 1989
  • Polysomal polyadenylated mRNAs which were purified from pea leaves were fractionated by sucrose grandient sedimentation. Fractions corresponding to the peak at 11.5S were found to contain mostly mRNA encoding the precursor polypeptide to the small subunit of ribulose bisphosphate carboxylase (rbcS) by in vitro translation in wheat germ extract. Double-stranded cDNA which was synthesized from the 11.5S mRNA was cloned into Hind III site of plasmid pBR 325. A cDNA clone, H24, was identified to code for rbcS. In vitro translation product of the hybridization-selected mRNA was molecular weight 20,000, presumably the precursor of rbcS. The nucleotide sequences of the H24 showed almost complete homology with the sequences encoding the transit peptide of the rbcS-3A gene which was reported by Fluhr et al.(1986).

  • PDF

Mechanism of guanine-specific DNA damage by UVA and its role in photocarcinogenesis and photoaging

  • Kawanishi, Shosuke;Oikawa, Shinji;Hiraku, Yusuke
    • Journal of Photoscience
    • /
    • 제9권2호
    • /
    • pp.150-153
    • /
    • 2002
  • Solar UV light is a well-known carcinogen. UVA radiation is probably carcinogenic to humans. In addition, recent investigations point to the importance of UVA irradiation in the photoaging. We investigated the mechanism of sequence- specific DNA damage using $\^$32/P-Iabeled DNA fragments in relation to carcinogenesis and aging. Furthermore, we investigated whether UVA accelerates the telomere shortening in human WI-38 fibroblasts. The exposure of double- stranded DNA fragments to 365 nm light in the presence of endogenous sensitizers produced sequence-specific cleavage at the 5' site of 5'-GG-3' and 5'-GGG-3' sequences. In addition, HPLC analysis revealed that sensitizers plus 365 nm light increased the 8-oxodG content of double-stranded DNA. We discuss the mechanisms of guanine-specific DNA damagecaused by excited photosensitizers in relation to carcinogenesis and aging.

  • PDF

Effective Family Shuffling Method Using Complementary DNA Fragments Produced by S1 Nuclease

  • Hong, Soon-Gyu
    • Journal of Microbiology and Biotechnology
    • /
    • 제16권12호
    • /
    • pp.2004-2007
    • /
    • 2006
  • An efficient method for the in vitro reassembly of homologous DNA sequences is presented. The proposed method involves obtaining single strands of homologous genes and hybridizing them to obtain partially hybridized heteroduplex DNA; cleaving the single-stranded regions of the heteroduplex DNA using S1 nuclease to generate double-strand DNA fragments; denaturing the double-strand DNA fragments to generate single-strand DNA fragments; conducting a series of polymerase chain reactions (PCR) using the single-strand DNA fragments as internal primers and a mixture of homologous DNA as templates to obtain elongated reassembled DNA; and finally, amplifying the reassembled DNA by a PCR using terminal primers. As a result, DNA reassembly could be achieved between homologous genes with a sequence similarity as low as 78%.

실력이 순서화된 경우에 대한 복식조 편성방법 (Method of Deciding Optimal Double Pairs When Players are Ordered)

  • 조대현
    • 응용통계연구
    • /
    • 제22권6호
    • /
    • pp.1331-1343
    • /
    • 2009
  • 본 논문에서는 테니스 동호인의 복식경기에 대하여 참여선수들의 실력이 순서화되어 있는 경우 공정하면서 각 경기가 가능한 대등한 경기가 될 수 있도록 복식조를 편성하는 방법을 다루었다. 라운드 로빙 방식과 파트너-상대 테이블과 ox(경기-휴식)표를 이용하여 경기 참가자수 및 게임 수에 따라 편성된 복식조를 이용하여 새롭게 정의된 손실함수 값을 최소로 하는 최적의 복식조를 결정하였다.

STRONG LAWS OF LARGE NUMBERS FOR WEIGHTED SUMS OF NEGATIVELY DEPENDENT RANDOM VARIABLES

  • Ko, Mi-Hwa;Han, Kwang-Hee;Kim, Tae-Sung
    • 대한수학회지
    • /
    • 제43권6호
    • /
    • pp.1325-1338
    • /
    • 2006
  • For double arrays of constants ${a_{ni},\;1{\leq}i{\leq}k_n,\;n{\geq}1}$ and sequences of negatively orthant dependent random variables ${X_n,\;n{\geq}1}$, the conditions for strong law of large number of ${\sum}^{k_n}_{i=1}a_{ni}X_i$ are given. Both cases $k_n{\uparrow}{\infty}\;and\;k_n={\infty}$ are treated.