Accumulated low density parity check （LDPCA） codec is proposed for DISCOVER project in distributed video coding （DVC）, which offers flexible coding rate. Although it can use feedback channel to adapt the rate to t...Accumulated low density parity check （LDPCA） codec is proposed for DISCOVER project in distributed video coding （DVC）, which offers flexible coding rate. Although it can use feedback channel to adapt the rate to the correlation of the video, but in real applications, using feedback channel can not always be possible. To solve this problem, some researchers proposed estimating the code rate at the encoder but the performance was not very good. Based on their researches, this paper considers the impact of convergence rate for iteration on rate estimate, which can be calculated using its check matrix. As a pilot study, this paper pays attention to the regular LDPCA codec. At the same time, it considers the impact of deviation in the estimated crossover probability, which gives some constraints to rate estimate. In the experiment, the proposed algorithm can improve the rate-distortion performance by up to 1 dB-1.2 dB.展开更多
In this paper, we obtain functional limit theorems for d-dimensional FBM in HSlder norm via estimating large deviation probabilities for d-dimensional FBM in HSlder norm.
Let {ξ<SUB> j </SUB>; j ∈ ℤ<SUB>+</SUB><SUP> d </SUP>be a centered stationary Gaussian random field, where ℤ<SUB>+</SUB><SUP>...Let {ξ<SUB> j </SUB>; j ∈ ℤ<SUB>+</SUB><SUP> d </SUP>be a centered stationary Gaussian random field, where ℤ<SUB>+</SUB><SUP> d </SUP>is the d-dimensional lattice of all points in d-dimensional Euclidean space ℝ<SUP>d</SUP>, having nonnegative integer coordinates. For each j = (j <SUB>1 </SUB>, ..., jd) in ℤ<SUB>+</SUB><SUP> d </SUP>, we denote |j| = j <SUB>1 </SUB>... j <SUB>d </SUB>and for m, n ∈ ℤ<SUB>+</SUB><SUP> d </SUP>, define S(m, n] = Σ<SUB> m【j≤n </SUB>ζ<SUB> j </SUB>, σ<SUP>2</SUP>(|n−m|) = ES <SUP>2 </SUP>(m, n], S <SUB>n </SUB>= S(0, n] and S <SUB>0 </SUB>= 0. Assume that σ(|n|) can be extended to a continuous function σ(t) of t 】 0, which is nondecreasing and regularly varying with exponent α at b ≥ 0 for some 0 【 α 【 1. Under some additional conditions, we study limsup results for increments of partial sum processes and prove as well the law of the iterated logarithm for such partial sum processes.展开更多
General limit theorems are established for l^p-valued Gaussian random fields indexed by a multidimensional parameter,which contain both almost sure moduli of continuity and limits of large increments for the l^p-value...General limit theorems are established for l^p-valued Gaussian random fields indexed by a multidimensional parameter,which contain both almost sure moduli of continuity and limits of large increments for the l^p-valued Gaussian random fields under(?)explicit conditions.展开更多
文摘Accumulated low density parity check （LDPCA） codec is proposed for DISCOVER project in distributed video coding （DVC）, which offers flexible coding rate. Although it can use feedback channel to adapt the rate to the correlation of the video, but in real applications, using feedback channel can not always be possible. To solve this problem, some researchers proposed estimating the code rate at the encoder but the performance was not very good. Based on their researches, this paper considers the impact of convergence rate for iteration on rate estimate, which can be calculated using its check matrix. As a pilot study, this paper pays attention to the regular LDPCA codec. At the same time, it considers the impact of deviation in the estimated crossover probability, which gives some constraints to rate estimate. In the experiment, the proposed algorithm can improve the rate-distortion performance by up to 1 dB-1.2 dB.
基金1)This work is supported by NSFC(10571159),SRFDP(2002335090)and KRF(D00008)2)This work is supported by NSFC(10401037)and China Postdoctoral Science Foundation3)This work is supported by the Brain Korea 21 Project in 2005
文摘In this paper, we obtain functional limit theorems for d-dimensional FBM in HSlder norm via estimating large deviation probabilities for d-dimensional FBM in HSlder norm.
基金NSERC Canada grants of Miklos Csorgo and Barbara Szyszkowicz at Carleton University,Ottawa,and by KRF-2003-C00098NSERC Canada grants at Carleton University,Ottawa
文摘Let {ξ<SUB> j </SUB>; j ∈ ℤ<SUB>+</SUB><SUP> d </SUP>be a centered stationary Gaussian random field, where ℤ<SUB>+</SUB><SUP> d </SUP>is the d-dimensional lattice of all points in d-dimensional Euclidean space ℝ<SUP>d</SUP>, having nonnegative integer coordinates. For each j = (j <SUB>1 </SUB>, ..., jd) in ℤ<SUB>+</SUB><SUP> d </SUP>, we denote |j| = j <SUB>1 </SUB>... j <SUB>d </SUB>and for m, n ∈ ℤ<SUB>+</SUB><SUP> d </SUP>, define S(m, n] = Σ<SUB> m【j≤n </SUB>ζ<SUB> j </SUB>, σ<SUP>2</SUP>(|n−m|) = ES <SUP>2 </SUP>(m, n], S <SUB>n </SUB>= S(0, n] and S <SUB>0 </SUB>= 0. Assume that σ(|n|) can be extended to a continuous function σ(t) of t 】 0, which is nondecreasing and regularly varying with exponent α at b ≥ 0 for some 0 【 α 【 1. Under some additional conditions, we study limsup results for increments of partial sum processes and prove as well the law of the iterated logarithm for such partial sum processes.
基金This work was supported by NSERC Canada grants at Carleton University and by KOSEF-R01-2005-000-10696-0
文摘General limit theorems are established for l^p-valued Gaussian random fields indexed by a multidimensional parameter,which contain both almost sure moduli of continuity and limits of large increments for the l^p-valued Gaussian random fields under(?)explicit conditions.
