Accuracy of Video Frame Size Forecasting

Micheal P. McGarry; Patrick Seeling; Rami J. Haddad; Jesus Hernandez

doi:10.1109/EIT.2012.6220754

Accuracy of Video Frame Size Forecasting

Micheal P. McGarry, Patrick Seeling, Rami J. Haddad, Jesus Hernandez

Electrical & Computer Engineering

Research output: Contribution to conference › Presentation

Abstract

We experimentally explore the forecast accuracy of three forecast model types, namely (i) Auto-Regressive (AR), (ii) Moving Average (MA), and (iii) Exponentially Weighted Moving Average (EWMA). Specifically, we use a set of five MPEG-4 video traces and measure the Normalized Mean Squared Error (NMSE) for each of the model types with varying model parameters. We find that increasing model order is not beneficial for I and P frame size forecasting, but is beneficial to B frame size forecasting. Most importantly, we find that the forecast accuracy has the strongest dependence on the statistical properties of the video content itself. It appears that the forecast accuracy is not dependent on the forecast model properties but rather strictly dependent on the logical and statistical properties of the content.

Original language	American English
DOIs	https://doi.org/10.1109/EIT.2012.6220754
State	Published - May 6 2012
Event	IEEE International Conference on Electro/Information Technology (EIT) - Duration: May 6 2012 → …

Conference

Conference	IEEE International Conference on Electro/Information Technology (EIT)
Period	05/6/12 → …

Keywords

Video frame size forecasting
Video frame size prediction

DC Disciplines

Electrical and Computer Engineering

Access to Document

10.1109/EIT.2012.6220754License: Unspecified

http://dx.doi.org/10.1109/EIT.2012.6220754

Cite this

@conference{7dc0bdc5abe04c8aa857eb6151d2da64,

title = "Accuracy of Video Frame Size Forecasting",

abstract = " We experimentally explore the forecast accuracy of three forecast model types, namely (i) Auto-Regressive (AR), (ii) Moving Average (MA), and (iii) Exponentially Weighted Moving Average (EWMA). Specifically, we use a set of five MPEG-4 video traces and measure the Normalized Mean Squared Error (NMSE) for each of the model types with varying model parameters. We find that increasing model order is not beneficial for I and P frame size forecasting, but is beneficial to B frame size forecasting. Most importantly, we find that the forecast accuracy has the strongest dependence on the statistical properties of the video content itself. It appears that the forecast accuracy is not dependent on the forecast model properties but rather strictly dependent on the logical and statistical properties of the content.",

keywords = "Video frame size forecasting, Video frame size prediction",

author = "McGarry, {Micheal P.} and Patrick Seeling and Haddad, {Rami J.} and Jesus Hernandez",

year = "2012",

month = may,

day = "6",

doi = "10.1109/EIT.2012.6220754",

language = "American English",

note = "IEEE International Conference on Electro/Information Technology (EIT) ; Conference date: 06-05-2012",

}

TY - CONF

T1 - Accuracy of Video Frame Size Forecasting

AU - McGarry, Micheal P.

AU - Seeling, Patrick

AU - Haddad, Rami J.

AU - Hernandez, Jesus

PY - 2012/5/6

Y1 - 2012/5/6

N2 - We experimentally explore the forecast accuracy of three forecast model types, namely (i) Auto-Regressive (AR), (ii) Moving Average (MA), and (iii) Exponentially Weighted Moving Average (EWMA). Specifically, we use a set of five MPEG-4 video traces and measure the Normalized Mean Squared Error (NMSE) for each of the model types with varying model parameters. We find that increasing model order is not beneficial for I and P frame size forecasting, but is beneficial to B frame size forecasting. Most importantly, we find that the forecast accuracy has the strongest dependence on the statistical properties of the video content itself. It appears that the forecast accuracy is not dependent on the forecast model properties but rather strictly dependent on the logical and statistical properties of the content.

AB - We experimentally explore the forecast accuracy of three forecast model types, namely (i) Auto-Regressive (AR), (ii) Moving Average (MA), and (iii) Exponentially Weighted Moving Average (EWMA). Specifically, we use a set of five MPEG-4 video traces and measure the Normalized Mean Squared Error (NMSE) for each of the model types with varying model parameters. We find that increasing model order is not beneficial for I and P frame size forecasting, but is beneficial to B frame size forecasting. Most importantly, we find that the forecast accuracy has the strongest dependence on the statistical properties of the video content itself. It appears that the forecast accuracy is not dependent on the forecast model properties but rather strictly dependent on the logical and statistical properties of the content.

KW - Video frame size forecasting

KW - Video frame size prediction

UR - https://digitalcommons.georgiasouthern.edu/electrical-eng-facpubs/40

UR - http://dx.doi.org/10.1109/EIT.2012.6220754

U2 - 10.1109/EIT.2012.6220754

DO - 10.1109/EIT.2012.6220754

M3 - Presentation

T2 - IEEE International Conference on Electro/Information Technology (EIT)

Y2 - 6 May 2012

ER -

Accuracy of Video Frame Size Forecasting

Abstract

Conference

Keywords

DC Disciplines

Access to Document

Other files and links

Fingerprint

Cite this