Skip to main content

Lund University Publications

LUND UNIVERSITY LIBRARIES

Towards classification of head movements in audiovisual recordings of read news

Frid, Johan LU orcid ; Ambrazaitis, Gilbert LU ; Svensson Lundmark, Malin LU orcid and House, David (2017) 4th European and 7th Nordic Symposium on Multimodal Communication In Linköping Electronic Conference Proceedings p.4-9
Abstract
In this paper we develop a system for detection of word-related head movements in audiovisu-al recordings of read news. Our materials consist of Swedish television news broadcasts and comprise audiovisual recordings of five news readers (two female, three male). The corpus was manually labelled for head movement, applying a simplistic annotation scheme consisting of a binary decision about absence/presence of a movement in relation to a word. We use OpenCV for frontal face detection and based on this we calculate velocity and acceleration features. Then we train a machine learning system to predict absence or presence of head movement and achieve an accuracy of 0.892, which is better than the baseline. The system may thus be helpful for... (More)
In this paper we develop a system for detection of word-related head movements in audiovisu-al recordings of read news. Our materials consist of Swedish television news broadcasts and comprise audiovisual recordings of five news readers (two female, three male). The corpus was manually labelled for head movement, applying a simplistic annotation scheme consisting of a binary decision about absence/presence of a movement in relation to a word. We use OpenCV for frontal face detection and based on this we calculate velocity and acceleration features. Then we train a machine learning system to predict absence or presence of head movement and achieve an accuracy of 0.892, which is better than the baseline. The system may thus be helpful for head movement labelling. (Less)
Abstract (Swedish)
In this paper we develop a system for detection of word-related head movements in audiovisu-al recordings of read news. Our materials consist of Swedish television news broadcasts and comprise audiovisual recordings of five news readers (two female, three male). The corpus was manually labelled for head movement, applying a simplistic annotation scheme consisting of a binary decision about absence/presence of a movement in relation to a word. We use OpenCV for frontal face detection and based on this we calculate velocity and acceleration features. Then we train a machine learning system to predict absence or presence of head movement and achieve an accuracy of 0.892, which is better than the baseline. The system may thus be helpful for... (More)
In this paper we develop a system for detection of word-related head movements in audiovisu-al recordings of read news. Our materials consist of Swedish television news broadcasts and comprise audiovisual recordings of five news readers (two female, three male). The corpus was manually labelled for head movement, applying a simplistic annotation scheme consisting of a binary decision about absence/presence of a movement in relation to a word. We use OpenCV for frontal face detection and based on this we calculate velocity and acceleration features. Then we train a machine learning system to predict absence or presence of head movement and achieve an accuracy of 0.892, which is better than the baseline. The system may thus be helpful for head movement labelling. (Less)
Please use this url to cite or link to this publication:
author
; ; and
organization
publishing date
type
Chapter in Book/Report/Conference proceeding
publication status
published
subject
host publication
Proceedings of the 4th European and 7th Nordic Symposium on Multimodal Communication (MMSYM 2016)
series title
Linköping Electronic Conference Proceedings
editor
Paggio, Patrizia and Navarretta, Costanza
article number
002
pages
6 pages
publisher
Linköping University Electronic Press
conference name
4th European and 7th Nordic Symposium on Multimodal Communication
conference location
Copenhagen, Denmark
conference dates
2016-09-29 - 2016-09-30
ISSN
1650-3686
1650-3740
ISBN
978-91-7685-423-5
project
SWE-CLARIN: Svensk språkteknologi för humaniora och samhällsvetenskap
Multi-modal levels of prominence: How verbal and visual signals interact in the coding of fine distinctions in information structure
language
English
LU publication?
yes
id
1c99e24d-d86a-499d-af65-0ea2006b236c
alternative location
http://www.ep.liu.se/ecp/141/002/ecp17141002.pdf
date added to LUP
2017-09-25 14:32:23
date last changed
2023-06-08 02:50:48
@inproceedings{1c99e24d-d86a-499d-af65-0ea2006b236c,
  abstract     = {{In this paper we develop a system for detection of word-related head movements in audiovisu-al recordings of read news. Our materials consist of Swedish television news broadcasts and comprise audiovisual recordings of five news readers (two female, three male). The corpus was manually labelled for head movement, applying a simplistic annotation scheme consisting of a binary decision about absence/presence of a movement in relation to a word. We use OpenCV for frontal face detection and based on this we calculate velocity and acceleration features. Then we train a machine learning system to predict absence or presence of head movement and achieve an accuracy of 0.892, which is better than the baseline. The system may thus be helpful for head movement labelling.}},
  author       = {{Frid, Johan and Ambrazaitis, Gilbert and Svensson Lundmark, Malin and House, David}},
  booktitle    = {{Proceedings of the 4th European and 7th Nordic Symposium on Multimodal Communication (MMSYM 2016)}},
  editor       = {{Paggio, Patrizia and Navarretta, Costanza}},
  isbn         = {{978-91-7685-423-5}},
  issn         = {{1650-3686}},
  language     = {{eng}},
  month        = {{09}},
  pages        = {{4--9}},
  publisher    = {{Linköping University Electronic Press}},
  series       = {{Linköping Electronic Conference Proceedings}},
  title        = {{Towards classification of head movements in audiovisual recordings of read news}},
  url          = {{http://www.ep.liu.se/ecp/141/002/ecp17141002.pdf}},
  year         = {{2017}},
}