Audio-Visual Emotion Recognition Based on a DBN Model with Constrained Asynchrony

Chen, Danqi; Jiang, Dongmei; Ravyse, Ilse; Sahli, Hichem

doi:10.1109/icig.2009.120

Cited by 8 publications

(12 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…7 shows an example of the tracking results. Experimental results demonstrate that the BTSM based tracking system can reliably and precisely track the face shape through long sequences [7].…”

Section: B Feature Extractionmentioning

confidence: 99%

See 1 more Smart Citation

Visual emotion recognition based on dynamic models

Fan

et al. 2013

2013 IEEE International Conference on Signal Processing, Communication and Computing (ICSPCC 2013)

View full text Add to dashboard Cite

This paper introduces the semi-continuous Hidden Markov Model (HMM) and proposes a novel Dynamic Bayesian Network (DBN) model for dynamic visual emotion recognition. The former reduces the training complexity caused by mixture Gaussians by sharing the Condition Probability Densities (CPDs) of Gaussians among the states, and the latter adds a sub-state layer between state and observation layer based on traditional DBN framework and describes the dynamic process of visual emotion in detail. Experiments results show that semi-continuous HMM and three-layer DBN have better performance, and average emotion recognition rate of the semi-continuous HMM is 1.85% and 3.82% higher than those of classical HMM and mixture Gaussian HMM respectively, and average emotion recognition rate of three-layer DBN is 1.93% higher than that of traditional DBN.

show abstract

Section: B Feature Extractionmentioning

confidence: 99%

“…Classical DBN model is shown as Fig.4 [7], which can be used for speech recognition and speaker recognition.…”

Section: B Traditional Dbn Modelmentioning

confidence: 99%

Visual emotion recognition based on dynamic models

Fan

et al. 2013

2013 IEEE International Conference on Signal Processing, Communication and Computing (ICSPCC 2013)

View full text Add to dashboard Cite

show abstract

“…In [13], a two stream state asynchronous DBN model (Asy_DBN) has been proposed. In this paper, to combine the MFCC features, local prosodic features and visual emotion features more reasonably, we extend the Asy_DBN model to a triple stream audio visual asynchronous DBN (T_AsyDBN) model, as shown in Fig.…”

Section: Triple Stream Asynchronous Dbn (T_asydbn) Modelmentioning

confidence: 99%

“…Fig.1 In [13], an 8-dimensional feature vector, describing the movements of eyes and eyebrows in the upper face, has been used. For the sake of completeness, these features are summarized in Table 1, where point 84 has been defined as the midpoint between landmarks 41 and 43, representing the nose apex.…”

Section: D Facial Featuresmentioning

confidence: 99%

“…To model the asynchrony in a more flexible way, it has been proposed in [13] a two stream dynamic Bayesian network model (Asy_DBN) using as audio inputs the perceptual linear prediction (PLP) coefficients, and as visual inputs facial shape features. This model allows the audio-state and visual-state to be asynchronous but within certain constraints.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Audio Visual Emotion Recognition Based on Triple-Stream Dynamic Bayesian Network Models

Jiang

Cui

Zhang

et al. 2011

Affective Computing and Intelligent Interaction

Self Cite

View full text Add to dashboard Cite

Abstract. We present a triple stream DBN model (T_AsyDBN) for audio visual emotion recognition, in which the two audio feature streams are synchronous, while they are asynchronous with the visual feature stream within controllable constraints. MFCC features and the principle component analysis (PCA) coefficients of local prosodic features are used for the audio streams. For the visual stream, 2D facial features as well 3D facial animation unit features are defined and concatenated, and the feature dimensions are reduced by PCA. Emotion recognition experiments on the eNERFACE'05 database show that by adjusting the asynchrony constraint, the proposed T_AsyDBN model obtains 18.73% higher correction rate than the traditional multi-stream state synchronous HMM (MSHMM), and 10.21% higher than the two stream asynchronous DBN model (Asy_DBN).

show abstract

Recognizing affect from speech prosody using hierarchical graphical models

Fernandez

Picard²

2011

Speech Communication

View full text Add to dashboard Cite

Audio-Visual Emotion Recognition Based on a DBN Model with Constrained Asynchrony

Cited by 8 publications

References 19 publications

Visual emotion recognition based on dynamic models

Visual emotion recognition based on dynamic models

Audio Visual Emotion Recognition Based on Triple-Stream Dynamic Bayesian Network Models

Recognizing affect from speech prosody using hierarchical graphical models

Contact Info

Product

Resources

About