Shoe designing support device that estimates impression index, and shoe designing support method that estimates impression index
A design element index input unit 10 inputs multiple design element indexes obtained by quantifying design in association with multiple portions constituting shoes. A learning unit 40 uses, as teacher data, pairs of multiple design element indexes and an impression index to learn correlations between multiple design element indexes and the impression index, and creates an impression estimation model. By means of a learned impression estimation model, an impression index estimation unit 60 estimates an impression index obtained by quantifying an impression of shoes, based on a set of multiple design element indexes that have been input. An output unit 70 outputs the estimated impression index.
1. A designing support device, comprising:
an input unit that inputs a plurality of design element indexes obtained by quantifying design in association with a plurality of portions constituting shoes;
a target segment input unit that inputs a plurality of target segments for evaluating a subject shoe;
a learning unit that receives the plurality of design element indexes input through the input unit and the plurality of target segments input through the target segment input unit, performs machine learning to obtain correlations between the plurality of the design element indexes and impression indexes of the shoes using the plurality of design element indexes and the plurality of target segments, and creates an impression estimation model based on the correlations obtained by the machine learning;
an estimation unit that estimates an impression index obtained by quantifying an impression of the shoes by using the created impression estimation model, based on a set of a plurality of the design element indexes that have been input; and
an output unit that outputs the impression index thus estimated,
wherein the design element indexes include at least one of a radius of curvature of an instep of the shoes, a tiptoe elevation angle of the shoes, or a height of the shoes.
2. The designing support device of claim 1 , wherein
the learning unit uses, as teacher data, pairs of a plurality of the design element indexes and the impression index for each of a plurality of segments into which users of the shoes are classified to learn the correlations between a plurality of the design element indexes and the impression index, and creates the impression estimation model for each of the segments,
the input unit further inputs a segment, and,
by means of the impression estimation model learned for each of the segments, the estimation unit estimates the impression index based on the segment and a plurality of the design element indexes that have been input.
3. The designing support device of claim 1 , wherein
the estimation unit extracts at least one design element index in descending order of a degree of influence on the impression index, and
the output unit outputs the design element index extracted for each impression index, together with the impression index.
4. The designing support device of claim 3 , wherein, with regard to the design element index extracted for each impression index, the output unit also outputs a direction of modification of the design element index for improving the impression index.
5. The designing support device of claim 3 , wherein, with regard to the design element index extracted for each impression index, the output unit also outputs a range of modification of the design element index for improving the impression index.
6. The designing support device of claim 3 , further comprising a design element index modification unit that modifies the design element index extracted for each impression index, wherein
the estimation unit re-estimates the impression index using the design element index thus modified.
7. The designing support device of claim 1 , wherein the design element indexes further include an index related to a color or texture in association with each of the plurality of portions constituting the shoes.
8. The designing support device of claim 1 , further comprising a recommendation unit that recommends other shoes having an impression index similar to the impression index that has been estimated.
9. A designing support method, comprising:
inputting a plurality of design element indexes obtained by quantifying design in association with a plurality of portions constituting shoes;
inputting a plurality of target segments for evaluating a subject shoe;
receiving the plurality of design element indexes and the plurality of target segments, performing machine learning to obtain correlations between the plurality of the design element indexes and impression indexes of the shoes using the plurality of design element indexes and the plurality of target segments, and creating an impression estimation model based on the correlations obtained by the machine learning;
estimating an impression index obtained by quantifying an impression of the shoes by using the created impression estimation model, based on a set of a plurality of the design element indexes that have been input; and
outputting the impression index thus estimated,
wherein the design element indexes include at least one of a radius of curvature of an instep of the shoes, a tiptoe elevation angle of the shoes, or a height of the shoes.