Audio-Visual Emotion Recognition Based on a DBN Model with Constrained Asynchrony
This paper presents an audio visual multi-stream DBn model (Asy_DBN) for emotion recognition with constraint asynchrony, in which audio state and visual state transit individually in their corresponding stream but the transition is constrained by the allowed maximum audio visual asynchrony. Emotion recognition experiments of Asy_DBN with different asynchrony constraints are carried out on an audio visual speech database of four emotions, and compared with the single streaM HMM, state synchronous HMM (Syn_HMM) and state synchronous DBN model, as well the state asynchronous DBN model without asynchrony constraint. Results show that by setting the appropriate maximum asynchrony constraint between audio and visual streams, the proposed audio visual asynchronous DBN model gets the highest emotion recognition performance, with an improvement of 15% over SynHMM.
audio visual multi-stream asynchronous DBN model
Danqi Chen Dongmei Jiang Use Ravyse Hichem Sahli
VUB-NPU Joint Research Group on AVSP Northwestern Polytechnic University, Xian, China Shaanxi Provi VUB-NPU Joint Research Group on AVSP Vrije Universiteit Brussel (VUB) - Interdisciplinary Institute
国际会议
The Fifth International Conference on Image and Graphics(第五届国际图像图形学学术会议 ICIG 2009)
西安
英文
912-916
2009-09-20（万方平台首次上网日期，不代表论文的发表时间）