Sorry, you need to enable JavaScript to visit this website.

facebooktwittermailshare

A PLLR and Multi-stage Staircase Regression Framework for Speech-based Emotion Prediction

Abstract: 

Continuous prediction of dimensional emotions (e.g. arousal and valence) has attracted increasing research interest recently. When processing emotional speech signals, phonetic features have been rarely used due to the assumption that phonetic variability is a confounding factor that degrades emotion recognition/prediction performance. In this paper, instead of eliminating phonetic variability, we investigated whether Phone Log-Likelihood Ratio (PLLR) features could be used to index arousal and valence in a pairwise low/high framework. A multi-stage staircase regression (SR) framework which enables fusion at three different stages is also investigated. Results on the RECOLA database show that PLLR outperforms EGEMAPS features for arousal and valence. Interestingly, long-term averaged PLLR proved to be more robust and emotionally informative than local frame-level PLLR, which contains more phoneme-specific information. Within the multi-stage SR framework, PLLR yielded an 8.2% and 11.6% relative improvement in CCC for arousal and valence respectively, showing great promise for including phonetic features in emotion prediction systems.

up
0 users have voted:

Paper Details

Authors:
Zhaocheng Huang, Julien Epps
Submitted On:
17 March 2017 - 10:17pm
Short Link:
Type:
Poster
Event:
Presenter's Name:
Zhaocheng Huang
Paper Code:
SP-P2.7
Document Year:
2017
Cite

Document Files

DAVID_ICASSP2017_V1.pdf

(37 downloads)

Subscribe

[1] Zhaocheng Huang, Julien Epps, "A PLLR and Multi-stage Staircase Regression Framework for Speech-based Emotion Prediction", IEEE SigPort, 2017. [Online]. Available: http://sigport.org/1775. Accessed: May. 30, 2017.
@article{1775-17,
url = {http://sigport.org/1775},
author = {Zhaocheng Huang; Julien Epps },
publisher = {IEEE SigPort},
title = {A PLLR and Multi-stage Staircase Regression Framework for Speech-based Emotion Prediction},
year = {2017} }
TY - EJOUR
T1 - A PLLR and Multi-stage Staircase Regression Framework for Speech-based Emotion Prediction
AU - Zhaocheng Huang; Julien Epps
PY - 2017
PB - IEEE SigPort
UR - http://sigport.org/1775
ER -
Zhaocheng Huang, Julien Epps. (2017). A PLLR and Multi-stage Staircase Regression Framework for Speech-based Emotion Prediction. IEEE SigPort. http://sigport.org/1775
Zhaocheng Huang, Julien Epps, 2017. A PLLR and Multi-stage Staircase Regression Framework for Speech-based Emotion Prediction. Available at: http://sigport.org/1775.
Zhaocheng Huang, Julien Epps. (2017). "A PLLR and Multi-stage Staircase Regression Framework for Speech-based Emotion Prediction." Web.
1. Zhaocheng Huang, Julien Epps. A PLLR and Multi-stage Staircase Regression Framework for Speech-based Emotion Prediction [Internet]. IEEE SigPort; 2017. Available from : http://sigport.org/1775