Predicting Image Emotion Distribution by Emotional Region

Fan, Yangyu; Yang, Hansen; Li, Zuhe; Liu, Shu

doi:10.1109/cisp-bmei.2018.8633190

Cited by 8 publications

(2 citation statements)

References 24 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Second, the closer the sample is to the predicted point, the greater its impact on the regression. Third, the sample point farther away from the regression line, the greater the error of regression prediction [39]. Motivated by those problems, this paper proposes the FWSVR model that adds the fuzzy membership function.…”

Section: The Framework Structure Of the Prediction Modelmentioning

confidence: 99%

Outlier Processing in Multimodal Emotion Recognition

et al. 2020

View full text Add to dashboard Cite

Automatic emotion recognition plays a key role in human-computer interactions. Multimodal emotion recognition has attracted much attention in recent years. When multimodalities are used, different modalities interact with each other and the obtained results tend to be accurate in general. However, there are also cases of unimodal anomalies. Most of the existing studies do not take into account the existence of outliers in the multimodality, which leads to low accuracy of the prediction results. This paper proposes fuzzy weighted support vector machine for regression (FWSVR) to deal with outliers and prediction errors. We design an automatic affective recognition model structure to analyze continuous dimension emotions based on multimodality (audio and visual). The LIRIS-ACCEDE database is used in this work. Experimental results indicate that the concordance correlation coefficient (CCC) is 0.9456 for arousal and 0.9183 for valence on the test set. The fusion result obtained when using fuzzy weighting is much better than the direct fusion one. INDEX TERMS Multimodal emotion recognition, fuzzy weighted support vector machine for regression (FWSVR), continuous dimension emotional space, outlier processing.

show abstract

Section: The Framework Structure Of the Prediction Modelmentioning

confidence: 99%

Outlier Processing in Multimodal Emotion Recognition

et al. 2020

View full text Add to dashboard Cite

show abstract

“…Kang and Yoon [29] constructed a multi-structure, variable-parameter expression identification model, which retains the sequential features of facial expressions, and solves the vanishing gradient problem caused by too many network layers. Fan et al [30] extracted the sequential features of facial expression images after blockbased preprocessing, quantified the associations between various facial expressions and emotions with the emotional index, and realized the continuous description of discrete facial expressions and effective recognition of the corresponding emotions. Balouchian and Foroosh [31] identified the micro-facial expressions by an end-to-end deep neural network, finetuned the parameters of the pretrained model through transfer learning to make up for the small size of sample set, and reduced the imbalance between different classes of micro-expression data with the focal loss function.…”

Section: Introductionmentioning

confidence: 99%

Machine Learning-Based Emotional Recognition in Surveillance Video Images in the Context of Smart City Safety

Pan¹,

Zhou²,

Liu³

et al. 2021

View full text Add to dashboard Cite

The effective extraction of deep information from surveillance video lays the basis for smart city safety. However, the surveillance video images contain complex targets, whose expression changes are difficult to capture. The traditional face expression recognition methods or sentiment analysis algorithms have a poor application effect. Based on machine learning (ML), this paper explores the emotional recognition in surveillance video images in the context of smart city safety. Firstly, the potential textures of surveillance video images were extracted under multi-order double cross (MODC) mode, and the optical flow features of facial expressions were detected in these images. Next, a facial expression recognition model was constructed based on the DeepID convolutional neural network (CNN), and an emotional semantic space was established for the face images in surveillance video. The proposed method was proved effective through experiments. The research results provide a reference for emotional recognition in images of other fields.

show abstract