Speech emotion recognition: Features and classification models

Chen, Lijiang; Mao, Xia; Xue, Yuli; Cheng, L. L.

doi:10.1016/j.dsp.2012.05.007

Cited by 208 publications

(82 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…• Speech Emotion Recognition: identify emotional states of human by analyzing their voice [8][9][10]. However, the main limitation of this approach is users need to speak in order to recognize their emotion.…”

Section: Emotion Recognitionmentioning

confidence: 99%

Healthcare System Focusing on Emotional Aspect using Augmented Reality

Tivatansakul

Ohkura

2014

Transactions of Japan Society of Kansei Engineering

View full text Add to dashboard Cite

Even though much recent research has proposed systems that provide assistance and services to people, these systems emphasize the support of the physical aspects at the expense of emotional aspects. However, emotional health is as important as physical health. Moreover, negative emotional health can lead to social or mental health problems. People with good emotional health can control themselves and can address their negative emotions. However, people with bad emotional health feel difficult to control and decrease their negative emotions. Therefore the negative emotional health can lead to social and mental health problems. To cope with negative emotional health in their daily life, we proposed a new healthcare system that focuses on emotional aspects. The system integrates augmented reality to display virtual objects in real environments and Kinect, which allows users to freely interact with them. We also employ biological sensors to measure and detect user emotions, and provide three services based on expected emotions: relaxation, amusement and excitement services. We designed and implemented a small prototype for our relaxation service called: a breathing control application. This application applies deep breathing techniques of stress management to supports users when they experience stress from society or work as well as other negative emotions. This application displays a virtual music box to assist them breathe deeply because virtual objects and music might increase user relaxation and decrease their stress. We also evaluated this application by performing the experiment. The stress and relaxation of the participants are measured using salivary amylase test. As the result, the application can decrease their stress.

show abstract

Section: Emotion Recognitionmentioning

confidence: 99%

Healthcare System Focusing on Emotional Aspect using Augmented Reality

Tivatansakul

Ohkura

2014

Transactions of Japan Society of Kansei Engineering

View full text Add to dashboard Cite

show abstract

“…Among all speech emotion features, prosodic feature and spectral feature are the most representative types of speech emotion features that are comprehensively employed in speech emotion recognition (Wu et al, 2011;Jin et al, 2013). The frequently adopted prosodic feature contains pitch, formants, energy, speed, and so on (Chen et al, 2012;Ayadi et al, 2011). The spectral feature is deemed to offer certain supplementary and various speech emotion features comparing with the prosodic feature and mel-frequency cepstral coefficients (MFCC), linear predictive cepstral coefficients (LPC) and log-frequency power coefficients (LFPC) are three most classic spectral features that are broadly adopted in a number of speech emotion recognition approaches (Wu et al, 2011;Ayadi et al, 2011;Jin et al, 2013).…”

Section: Introductionmentioning

confidence: 99%

“…How to pick up the most beneficial and serviceable speech emotion features which can accessibly and accurately represent the speech emotion information is the pivotal problem for speech emotion recognition (Chen et al, 2012;Jin et al, 2013;Ayadi et al, 2011). Among all speech emotion features, prosodic feature and spectral feature are the most representative types of speech emotion features that are comprehensively employed in speech emotion recognition (Wu et al, 2011;Jin et al, 2013).…”

Section: Introductionmentioning

confidence: 99%

Speech Emotion Recognition Based on Sparse Representation

Yan¹,

Wang²,

Gu³

et al. 2013

Archives of Acoustics

View full text Add to dashboard Cite

Speech emotion recognition is deemed to be a meaningful and intractable issue among a number of domains comprising sentiment analysis, computer science, pedagogy, and so on. In this study, we investigate speech emotion recognition based on sparse partial least squares regression (SPLSR) approach in depth. We make use of the sparse partial least squares regression method to implement the feature selection and dimensionality reduction on the whole acquired speech emotion features. By the means of exploiting the SPLSR method, the component parts of those redundant and meaningless speech emotion features are lessened to zero while those serviceable and informative speech emotion features are maintained and selected to the following classification step. A number of tests on Berlin database reveal that the recognition rate of the SPLSR method can reach up to 79.23% and is superior to other compared dimensionality reduction methods.

show abstract

“…In multiclass SVM classifier, for example, the structure of the classifier can be one-to-one, one-to-all, hierarchy, or tree structure, so several SVM nodes or models exist in the multiclass classifier [1][2][3]. There are two questions in speech emotion recognition (SER): (1) how to seek the optimal feature subset from the acoustic features; (2) whether the same acoustic feature subset is proper in all nodes of the multiclass classifier. These questions are researched in this paper.…”

Section: Introductionmentioning

confidence: 99%

Reordering Features with Weights Fusion in Multiclass and Multiple-Kernel Speech Emotion Recognition

Jiang

Xia

Wang

et al. 2017

Journal of Electrical and Computer Engineering

View full text Add to dashboard Cite

The selection of feature subset is a crucial aspect in speech emotion recognition problem. In this paper, a Reordering Features with Weights Fusion (RFWF) algorithm is proposed for selecting more effective and compact feature subset. The RFWF algorithm fuses the weights reflecting the relevance, complementarity, and redundancy between features and classes comprehensively and implements the reordering of features to construct feature subset with excellent emotional recognizability. A binary-tree structured multiple-kernel SVM classifier is adopted in emotion recognition. And different feature subsets are selected in different nodes of the classifier. The highest recognition accuracy of the five emotions in Berlin database is 90.549% with only 15 features selected by RFWF. The experimental results show the effectiveness of RFWF in building feature subset and the utilization of different feature subsets for specified emotions can improve the overall recognition performance.

show abstract

Speech emotion recognition: Features and classification models

Cited by 208 publications

References 16 publications

Healthcare System Focusing on Emotional Aspect using Augmented Reality

Healthcare System Focusing on Emotional Aspect using Augmented Reality

Speech Emotion Recognition Based on Sparse Representation

Reordering Features with Weights Fusion in Multiclass and Multiple-Kernel Speech Emotion Recognition

Contact Info

Product

Resources

About