Method, system and non-transitory computer-readable recording medium for analyzing fashion attributes of image data group using large image data

1 . A method for analyzing fashion attributes using large amounts of pieces of image data, comprising the steps of:

collecting image data including at least one item;

generating at least one album by dividing the image data based on at least one of a collection source and a collection period;

generating location information of a fashion item for each of the image data included in the at least one album by using an attribute classification AI model, and recognizing and labeling an attribute of the fashion item corresponding to the generated location information;

calculating, based on the labeled attributes, the number of items corresponding to each sub-item of each attribute among the image data included in the album and a proportion of each sub-item within the corresponding attribute; and

visualizing analysis results including the calculated number of items and the proportion of each sub-item within the corresponding attribute,

wherein the attribute classification AI model is a model for detecting the at least one item included in the image data, and for recognizing, labeling and classifying fashion attributes of the at least one item.

2 . The method of claim 1 , wherein in the step of collecting the image data, the image data is input and collected from a user terminal or a predetermined database.

3 . The method of claim 1 ,

wherein the attribute includes at least one of category, color, material, fit, print, style, shape, detail, neckline, and sleeve length.

4 . The method of claim 1 , wherein the step of calculating, based on the labeled attributes, the number of items corresponding to each sub-item of each attribute among the image data included in the album and a proportion of each sub-item within the corresponding attribute comprises performing the analysis at a predetermined interval and storing the analysis results.

5 . The method of claim 1 , wherein the step of visualizing analysis results including the calculated number of items and the proportion of each sub-item within the corresponding attribute comprises providing at least one of a number of items or a graph corresponding to a sub-item by attribute for the at least one item included in each of the image data.

6 . A non-transitory computer-readable recording medium having stored thereon a computer program for executing a method of claim 1 .

7 . A fashion attribute analysis server using large amounts of pieces of image data, comprising:

a memory storing at least one instruction; and

at least one processor configured to be connected with the memory, wherein the at least one processor executes the at least one instruction to:

collect image data including at least one item;

generate at least one album by dividing the image data based on at least one of a collection source and a collection period;

generate location information of a fashion item for each of the image data included in the at least one album by using an attribute classification AI model, and recognize and label an attribute of the fashion item corresponding to the generated location information;

calculate, based on the labeled attributes, the number of items corresponding to each sub-item of each attribute among the image data included in the at least one album and a proportion of each sub-item within the corresponding attribute; and

generate visualization data including the calculated number of items and the proportion of each sub-item within the corresponding attribute,

wherein the attribute classification AI model is a model for detecting the at least one item included in the image data and for recognizing, labeling and classifying fashion attributes of the at least one item.