Product color emotional design based on a convolutional neural network and search neural network

Abstract

Due to the growing trend of social manufacturing, product design has focused on meeting the emotional needs of users. As a product attribute, color plays an important role in meeting these needs. Therefore, product color emotional design has attracted the attention of researchers. However, a user's perception of the emotional image of product color is highly complex, and it is difficult to define this perception accurately. To this end, based on the theoretical framework of Kansei engineering, this study proposes a product color emotional design method based on a convolutional neural network and a search neural network. First, we implement a semantic differential experiment to ascertain the user's color image. Then we use a convolutional neural network to establish a complex association model between the product color and the user's emotional imagery. Based on this model, the search neural network is used to search and generate the product color design scheme that meets the target image. Finally, a product color design system applicable to practical design problems is developed. An example of the design of a home service robot demonstrates that the proposed method and system provides accurate product color design solutions that meet the needs of the user's emotional image and can be used to develop practical large‐scale applications of product color emotional design theory and methods.

An eye- tracking technology and MLP-based color matching design method

“…This indicates that it is indeed easy for people to look at the animated images they like in hearts, which is consistent with the findings of Xu. 13 However, different from Hsiao et al 19 and Ding et al 21 , the eye-tracking technology extracts a more intuitive heat zone map. Besides, the adoption of eye-tracking heat zone maps avoids the calculation of complex formulas such as color clustering, which reduces the requirement for expertise in color reuse of animated images and improves the efficiency of product color design 11 , 16 , 19 , 21 .…”

“…The other is to realize the mapping to the target product color scheme according to the image color semantics of plane images 14 . As is known to all, it is an important task to mine consumers' perception of colors, establish mapping relationship between consumers' emotions and colors, and further convert emotional images to product color design elements 14 , 21 , 22 .…”

Converting animated image colors into product colors: An eye- tracking technology and MLP-based color matching design method

“…This indicates that it is indeed easy for people to look at the animated images they like in hearts, which is consistent with the ndings of Xu. 13 However, different from Hsiao et al 19 and Ding et al 21 , the eye-tracking technology extracts a more intuitive heat zone map. Besides, the adoption of eye-tracking heat zone maps avoids the calculation of complex formulas such as color clustering, which reduces the requirement for expertise in color reuse of animated images and improves the e ciency of product color design 11,16,19,21 .…”

“…The MLP is the core of machine learning and deep learning, which has the advantages of self-adaptability, self-learning, real-time, high robustness and so on. Multiple hidden layers facilitate scholars to build powerful models for e ciently solving nonlinear system problems 21 . All the above studies used MIPs for screening and decision making of practical problems, and the experimental ndings demonstrated that the MIP has substantial improvement in stability and accuracy compared with other algorithms [37][38][39] .…”

Automatic coloration of pattern based on color parsing of Sung porcelain

“…Based on the research of Xu et al 14 and Sun et al, 15 Zhang et al 16 transferred the open innovation team collaboration mode to the design team collaboration environment for production color design tasks, and developed a color design simulation prototype system based on CorelDRAW. Some other researchers used some mainstream intelligent color matching algorithms, such as Interactive Genetic Algorithm (IGA), 17,18 Neural Network Model, 19,20 Artificial Bee Colony (ABC), 21 Evolutionary Algorithm (EA), 22 and Particle Swarm Optimization (PSO). 23 The above algorithm and color matching methods realize the overall image migration of scene color, but lack the recognition and free adjustment of the internal characteristics of the target, and are inconvenient to assist designers in color matching.…”