Publication Details

High Performance Architecture for Object Detection in Streamed Videos

ZEMČÍK, P.; JURÁNEK, R.; MUSIL, M.; MUSIL, P.; HRADIŠ, M. High Performance Architecture for Object Detection in Streamed Videos. Proceedings of FPL 2013. Porto: IEEE Circuits and Systems Society, 2013. p. 1-4. ISBN: 978-1-4799-0004-6.
Czech title
Architektura pro detekci objektů ve streamovaném videu
Type
conference paper
Language
English
Authors
Zemčík Pavel, prof. Dr. Ing., dr. h. c. (DCGM)
Juránek Roman, Ing., Ph.D. (DCGM)
Musil Martin, Ing., Ph.D. (DCGM)
Musil Petr, Ing., Ph.D. (DCGM)
Hradiš Michal, Ing., Ph.D. (DCGM)
URL
Keywords

Object Detection, FPGA, LRD, LBP, Streamed Video

Abstract

In this paper, we introduce a novel architecture of an engine for high performance multi-scale detection of objects in videos based on WaldBoost training algorithm.The key properties of the architecture include processing of streamed data and low resource consumption. We implemented the engine in FPGA and we show that it can process 640x480 pixel video streams at over 160 fps without the need of external memory. We evaluate the design on the face detection task, compare it to state of the art designs, and discuss its features and limitations.

Published
2013
Pages
1–4
Proceedings
Proceedings of FPL 2013
ISBN
978-1-4799-0004-6
Publisher
IEEE Circuits and Systems Society
Place
Porto
DOI
10.1109/FPL.2013.6645559
BibTeX 
@inproceedings{BUT103530,
  author="Pavel {Zemčík} and Roman {Juránek} and Martin {Musil} and Petr {Musil} and Michal {Hradiš}",
  title="High Performance Architecture for Object Detection in Streamed Videos",
  booktitle="Proceedings of FPL 2013",
  year="2013",
  pages="1--4",
  publisher="IEEE Circuits and Systems Society",
  address="Porto",
  doi="10.1109/FPL.2013.6645559",
  isbn="978-1-4799-0004-6",
  url="http://ieeexplore.ieee.org/xpl/articleDetails.jsp?tp=&arnumber=6645559"
}
Back to top