Font Size: a A A

Research On Emotion Model Of Lines And Derived Forms

Posted on:2022-07-07Degree:MasterType:Thesis
Country:ChinaCandidate:Y Q HuFull Text:PDF
GTID:2518306527482954Subject:Software engineering
Abstract/Summary:
Different line styles can convey different emotions.Exploring the emotional expression model of line and its derivative forms can reveal the fundamental factors that affect emotional expression,develop more emotional expression forms,and promote the development of emotional expression software and the emotional recognition of different expression forms by artificial intelligence.The work of this paper is divided into the following four aspects:1.Explore the mapping relationship between different line styles and emotions.In order to explore the emotional expression of different styles of lines,87 types of static horizontal lines were created.27 emotional valence words and 2 emotional arousal words were used.Testers were recruited to conduct online tests to match emotional words for each style of lines.Principal component analysis(PCA)was used to reduce the emotion dimension.It was found that the emotion space built by the first two-dimensional principal components had much in common with the classical 2D emotion model.By analyzing the distribution of different styles of lines in the two-dimensional emotional space,four mapping rules between style attributes and emotions were obtained.2.Explore the correlation between the line and its implied horizontal movement,and explore whether the mapping rules between the line and emotion can exist across forms.In order to explore the correlation between the line and its implied horizontal motion,the implied motion was visualized as a black spot with trailing effect,and 87 motion modes corresponding to the static line styles were created,including fast and slow speed.We recruited testers to do online tests and match emotional words for different samples.PCA was used to establish the emotional coordinate space for the collected data,and the distribution of static lines and dynamic samples in the emotional space was compared.Pearson correlation analysis was used to analyze the scores of 29 components of PCA.It was found that there was a strong correlation between the static line and its implied motion in emotional potency and emotional arousal.Further,we speculated whether the mapping relationship between styles and emotions can be derived to other forms.Constructed another two types of samples by program: 1)dialogue bubbles with static lines as the border;2)center vibration filtering the horizontal displacement of lines,which contains two values of velocity.Each form also created 87 styles for online testing.PCA was used to build a low dimensional emotional space.The distribution of four forms of samples in the emotional space was analyzed.Pearson correlation analysis was used to analyze the PCA component scores.The results shows that the mapping law of style and emotion will produce a certain deviation or enhancement when it extended to other forms.3.Verify whether the mapping law of dynamic center vibration form and emotion can be extended to other frequencies.It is known from the previous study that the mapping law between dynamic derivation form and emotion will shift with the frequency of motion.The paper further explored whether the mapping law changes linearly with the vibration frequency,and the influence of vibration phase and amplitude on the results.The test system was developed,the testers were recruited to do online tests,and the sensing frequencies of the testers to different frequencies were recorded.The results show that the sensing frequency and the real frequency are basically linear.Although the phase and amplitude will affect the sensing accuracy,the linear relationship is preserved under different conditions.That is,the mapping law of central vibration form and emotion is linear affected by the motion speed.4.Develop a multi-form emotion visualization system based on the emotion model of lines and derived forms.The system constructed a two-dimensional emotional space with the first two principal components obtained by PCA,showed the distribution of the tested samples in the emotional space,and provided three interaction types,realized the user to point the emotional coordinates in the emotional space,calculated the generation equation of the coordinate corresponding to the style by interpolation,and generated the sample corresponding to the emotional representation of the point.This paper gives the generation equations of the above 87 styles,the implementation details of the four forms,the design of empirical research and the process of data analysis,and describes the design and development of the emotional expression system.Finally,it makes a summary and prospect of the related work.
Keywords/Search Tags:Emotion mapping, line style, metaphor, principal component analysis, emotion visualization system
Related items