Analysis of optimal dynamic rate shaping of Markov-1 sources

Batra, Pankaj; Eleftheriadis, Alexandros

doi:10.1109/icip.2003.1247363

Cited by 4 publications

(7 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The importance of this result is that it allows us to use the much simpler memoryless DRS algorithms without any significant distortion penalty. For the sake of completeness, we repeat a part of the problem formulation from [4], [17]- [20] in the next section.…”

Section: B Contributionsmentioning

confidence: 99%

“…Thus, the minimization problem for the intercoded case becomes (4) Here, is the accumulated error due to motion compensation based on (optimal) truncation of DCT coefficients in the past frame, while is current-frame only distortion due to rate shaping of transformed prediction error. Note that in general , i.e., motion compensation is a nonlinear operation, because it involves integer arithmetic with truncation away from zero.…”

Section: Constrained Drs Of Intercoded Picturesmentioning

confidence: 99%

“…This section details the behavior of the rate shaping minimization problem for the motion compensated (inter) frame case assuming an autoregressive (AR) source model [4]. By analyzing the statistical and rate-distortion behavior of the different components of this minimization problem, the following key result is mathematically shown: the set of optimal breakpoint values for any frame is somewhat invariant to the accumulated motion compensated shaping error from past frames and may be very reasonably approximated using the current frame shaping error alone.…”

Section: Problem Analysis For Rate Shaping Of Ar(1) Sourcesmentioning

confidence: 99%

See 2 more Smart Citations

Dynamic rate shaping of compressed digital video

Eleftheriadis

Batra

2006

IEEE Trans. Multimedia

Self Cite

View full text Add to dashboard Cite

We discuss new theoretical and experimental results on the dynamic rate shaping (DRS) approach for transcoding compressed video bitstreams (MPEG-1, MPEG-2, MPEG-4, H.261, as well as JPEG). We analyze the behavior of DRS assuming a first order autoregressive source. We propose a set of low complexity algorithms for both constrained and unconstrained DRS and substantiate the almost-optimal experimental performance of the memoryless algorithm by assuming a first order autoregressive source. By deriving the statistical and rate-distortion characteristics of different components of the interframe rate shaping problem, we offer an explanation as to why the set of optimal breakpoint values for any frame is somewhat invariant to the accumulated motion compensated shaping error from past frames. We also present an extensive experimental study on the various DRS algorithms (causally optimal, memoryless, and rate-based) both in their constrained and generalized forms. The study proves the computational viability of the DRS approach to transcoding and identifies a range of rate shaping ratios for which it is better than requantization, both complexity-wise as well as in performance. This result is significant in that it opens up the way to construct much simpler memoryless algorithms that give minimal penalty in achieved quality, not just for this but possibly other types of algorithms. This is also the very first use of matrix perturbation theory for tracking the spectral behavior of the autocorrelation matrix of the source signal and the motion residual it yields.

show abstract

Section: B Contributionsmentioning

confidence: 99%

Section: Constrained Drs Of Intercoded Picturesmentioning

confidence: 99%

Section: Problem Analysis For Rate Shaping Of Ar(1) Sourcesmentioning

confidence: 99%

See 1 more Smart Citation

Dynamic rate shaping of compressed digital video

Eleftheriadis

Batra

2006

IEEE Trans. Multimedia

Self Cite

View full text Add to dashboard Cite

show abstract

“…Let x j be the decoded value of the m-th pixel in scan order within block y j,i , and let x j-1 be the motioncompensated component of x j , that is, the m-th value of block C j,i (y j-1 ) in the same scan order. If the relationship between these two values can be described as a first-order autoregressive [AR(1)] sequence [16] and random MSE distortion (0) j δ is added to x j-1 , the distortion is propagated to x j with (1) …”

Section: An Adaptive Estimation Model For the Overall CD Distortionsmentioning

confidence: 99%

“…Simulation Results for UDA Algorithm Figure 6 shows the simulation results for the Akiyo and Coast Guard sequences, which were controlled for MSE=15 and MSE=25, respectively. In order to compare the performance of the UDA algorithm, a uniform distortion-based CD (UDCD) scheme was chosen [1], [5], [6], [10], [12], [16], [17]. In the UDCD scheme, without considering the propagated errors of each CD distortion, DCT coefficients corresponding to a constant distortion per frame are dropped.…”

Section: P4_10%cd Sum_modelsmentioning

confidence: 99%

An Adaptive Control for the Propagation Errors Incurred by DCT Coefficient-Dropping Transcoder

Kim¹,

Seo²,

Yun³

2007

ETRI J

View full text Add to dashboard Cite

This paper presents a new distortion control scheme with a simple estimation model for the propagation errors incurred by dropping some parts of the bitstream in a frame dropping-coefficient dropping (FD-CD) transcoder. The primary goal of this paper is to facilitate bit-rate conversions and rate-distortion controls in the compressed domain without introducing a full decoding and reencoding system in the pixel domain. First, the error propagation behavior over several frame sequences due to coefficient dropping is investigated on the basis of statistical and empirical properties. Then, such properties are used to develop a simple estimation model for the CD distortion accounting for the characteristics of the underlying coded-frame. Finally, the proposed estimation model allows us to determine the amount of coefficient dropping and to effectively allocate rate-distortions into coded-frames. Experimental results show that the proposed estimation model accurately describes the characteristics of propagation errors adaptively in the compressed domain and can be easily applied to distortion control over different kinds of video sequences.Keywords: Coefficient dropping, propagation error, distortion control. Manuscript received Jan. 10, 2007; revised Aug. 14, 2007. Jin-soo Kim (phone: + 82 42 821 1147 and Mong-han Yun (email: monghan@pinetelecom.com I. IntroductionIn the framework of universal multimedia access (UMA), one challenge for video transmission is to deliver video content through heterogeneous network channels matching the diversity of client devices [1]. As one approach, video transcoding schemes are commonly used [2]. In particular, transcoding based on dropping some parts of the bitstream in the compressed domain is used mainly due to its low computational complexity and simple implementation [3], [4]. Since a frame memory is not required, and there is no need for an inverse discrete cosine transform (IDCT), regardless of the techniques used to achieve the reduced bit-rate, these schemes are relatively simple. This paper focuses on the frame dropping-coefficient dropping (FD-CD) scheme, which provides trade-offs between spatial and temporal qualities and extends the range of rate reduction. By combining framedropping and DCT CD, it is possible to simply adapt the bit rate of a pre-coded video to the available dynamic bandwidth, especially in streaming applications. Furthermore, the FD-CD scheme can achieve rate shaping across a wide range of rate reduction while maintaining spatial-temporal quality as much as possible. For a particular FD-CD operation, in which nonreference frames (namely, B frames) are dropped to avoid the decoding dependency, CD is applied to meet the available bandwidth quite accurately by adjusting the number of dropped coefficients.Some previous studies on FD-CD transcoders have focused on the optimal selection of the number of dropped coefficients within a frame [1], [5]- [9]. The dynamic rate shaping (DRS) approach was introduced as a technique to adapt the rate of compressed v...

show abstract

Efficient Control for the Distortion Incurred by Dropping DCT Coefficients in Compressed Domain

Kim

2005

Advances in Multimedia Information Processing - PCM 2005

View full text Add to dashboard Cite

Analysis of optimal dynamic rate shaping of Markov-1 sources

Cited by 4 publications

References 7 publications

Dynamic rate shaping of compressed digital video

Dynamic rate shaping of compressed digital video

An Adaptive Control for the Propagation Errors Incurred by DCT Coefficient-Dropping Transcoder

Efficient Control for the Distortion Incurred by Dropping DCT Coefficients in Compressed Domain

Contact Info

Product

Resources

About