Automatic Role Recognition Based on Conversational and Prosodic Behaviour
Salamin, Hugues and Truong, Khiet and Mohammadi, Gelareh (2010) Automatic Role Recognition Based on Conversational and Prosodic Behaviour. In: ACM International Conference on Multimedia, October 25-29, 2010, Firenze, Italy (pp. pp. 847-850).
Restricted to UT campus only : Request a copy
|Abstract:||This paper proposes an approach for the automatic recognition of roles in settings like news and talk-shows, where roles correspond to speciﬁc functions like Anchorman, Guest or Interview Participant. The approach is based on purely nonverbal vocal behavioral cues, including who talks when and how much (turn-taking behavior), and statistical properties of pitch, formants, energy and speaking rate (prosodic behavior). The experiments have been performed over a corpus of around 50 hours of broadcast material and the accuracy, percentage of time correctly labeled in terms of role, is up to 89%. Both turn-taking and prosodic behavior lead to satisfactory results. Furthermore, on one database, their combination leads to a statistically signiﬁcant improvement.
|Item Type:||Conference or Workshop Item|
Electrical Engineering, Mathematics and Computer Science (EEMCS)
|Link to this item:||http://purl.utwente.nl/publications/74618|
|Export this item as:||BibTeX|
Daily downloads in the past month
Monthly downloads in the past 12 months
Repository Staff Only: item control page