Publication detail

Spectrum-based approach to classifying video sequences for encoding experiments

SLANINA, M.

Original Title

Spectrum-based approach to classifying video sequences for encoding experiments

English Title

Spectrum-based approach to classifying video sequences for encoding experiments

Type

conference paper

Language

en

Original Abstract

With the evolution of video compression algorithms, transmission techniques as well as error concealment tools, benchmarking is essential to assess the performance of emerging techniques. This paper discusses the selection of test material in order to obtain a representative set of different scene character- istics. While the most common approach is relying on the Spatial Information and Temporal Information parameters, our results show that the spatio-temporal spectrum information can be used to classify sequences with high reliability while maintaining low computational complexity.

English abstract

With the evolution of video compression algorithms, transmission techniques as well as error concealment tools, benchmarking is essential to assess the performance of emerging techniques. This paper discusses the selection of test material in order to obtain a representative set of different scene character- istics. While the most common approach is relying on the Spatial Information and Temporal Information parameters, our results show that the spatio-temporal spectrum information can be used to classify sequences with high reliability while maintaining low computational complexity.

Keywords

Video compression, H.264, test procedure

Released

20.06.2018

ISBN

978-1-5386-6978-5

Book

2018 25th International Conference on Systems, Signals and Image Processing (IWSSIP)

Pages from

1

Pages to

5

Pages count

5

BibTex


@inproceedings{BUT148671,
  author="Martin {Slanina}",
  title="Spectrum-based approach to classifying video sequences for encoding experiments",
  annote="With the evolution of video compression algorithms, transmission techniques as well as error concealment tools, benchmarking is essential to assess the performance of emerging techniques. This paper discusses the selection of test material in order to obtain a representative set of different scene character- istics. While the most common approach is relying on the Spatial Information and Temporal Information parameters, our results show that the spatio-temporal spectrum information can be used to classify sequences with high reliability while maintaining low computational complexity.",
  booktitle="2018 25th International Conference on Systems, Signals and Image Processing (IWSSIP)",
  chapter="148671",
  doi="10.1109/IWSSIP.2018.8439344",
  howpublished="electronic, physical medium",
  year="2018",
  month="june",
  pages="1--5",
  type="conference paper"
}