[News] EX Kansei technology report has been released.

EX Kansei was featured in Technology Report Vol. 20, a collection of technical papers summarizing Konica Minolta's research and development published every year.

Please take a look at the detailed introduction and usage examples of each analysis technology implemented in EX Kansei.

 

Citation of abstract

We will introduce the technology that characterizes ``EXplainable Kansei,'' a new business based on the concept of ``Science of design that sells,'' which quantitatively analyzes design images to identify issues and propose improvements.

The technology installed can be divided into the following six items. 1. Attention analysis that predicts whether visual attention will be directed to appealing points; 2. Impression analysis to predict the impression a design image gives to an observer; 3. Automatic determination of whether color harmony theory conforms to the principle of order, 4. Automatic color harmonization that automatically harmonizes colors and presents improvement plans; 5. 6. Quantification of the degree of complexity representing the degree of crowding of the contour line; Automatic positioning map generation technology using principal component analysis. These technologies are based on academically supported algorithms with Konica Minolta's own improvements. Moreover, by packaging it according to the concept of EXplainable sensibility, it became a valuable software and service product. In order to continue providing value in the future, we will continue to improve each algorithm and dataset while considering the addition of new analysis methods and image features.

At the end of this article, we present an example of how the analysis technology we provide can improve sales by improving photos on a restaurant's reservation site.

 

You can check the report here.