Improved Emotion Recognition With a Novel Speaker-Independent Feature

Cited 38 time in webofscience Cited 0 time in scopus
  • Hit : 652
  • Download : 0
Emotion recognition is one of the latest challenges in human-robot interaction. This paper describes the realization of emotional interaction for a Thinking Robot, focusing on speech emotion recognition. In general, speaker-independent systems show a lower accuracy rate compared with speaker-dependent systems, as emotional feature values depend on the speaker and their gender. However, speaker-independent systems are required for commercial applications. In this paper, a novel speaker-independent feature, the ratio of a spectral flatness measure to a spectral center (RSS), with a small variation in speakers when constructing a speaker-independent system is proposed. Gender and emotion are hierarchically classified by using the proposed feature (RSS), pitch, energy, and the mel frequency cepstral coefficients. An average recognition rate of 57.2% (+/- 5.7%) at a 90% confidence interval is achieved with the proposed systems in the speaker-independent mode.
Publisher
IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
Issue Date
2009-06
Language
English
Article Type
Article
Keywords

HUMAN-COMPUTER INTERACTION; SPEECH RECOGNITION; SYSTEM; INTERFACE; STRESS; NOISE

Citation

IEEE-ASME TRANSACTIONS ON MECHATRONICS, v.14, no.3, pp.317 - 325

ISSN
1083-4435
DOI
10.1109/TMECH.2008.2008644
URI
http://hdl.handle.net/10203/96882
Appears in Collection
ME-Journal Papers(저널논문)
Files in This Item
There are no files associated with this item.
This item is cited by other documents in WoS
⊙ Detail Information in WoSⓡ Click to see webofscience_button
⊙ Cited 38 items in WoS Click to see citing articles in records_button

qr_code

  • mendeley

    citeulike


rss_1.0 rss_2.0 atom_1.0