WO2021009880A1 - デザイン支援装置及びデザイン支援方法 - Google Patents
デザイン支援装置及びデザイン支援方法 Download PDFInfo
- Publication number
- WO2021009880A1 WO2021009880A1 PCT/JP2019/028148 JP2019028148W WO2021009880A1 WO 2021009880 A1 WO2021009880 A1 WO 2021009880A1 JP 2019028148 W JP2019028148 W JP 2019028148W WO 2021009880 A1 WO2021009880 A1 WO 2021009880A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- index
- impression
- design element
- design
- shoe
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Ceased
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2111/00—Details relating to CAD techniques
- G06F2111/20—Configuration CAD, e.g. designing by assembling or positioning modules selected from libraries of predesigned modules
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2113/00—Details relating to the application field
- G06F2113/12—Cloth
Definitions
- the present invention relates to a technique for supporting product design and a technique for recommending a product.
- Patent Document 1 discloses a design construction support system that shows a more preferable alternative candidate from the difference calculation with the evaluation of the system while allowing the user to change the design.
- Patent Document 2 discloses an information processing device that generates an article design based on the content selected by the user and presents the generated article design to the user.
- Patent Document 3 discloses a method of determining the relative preference of one or more consumers from a group of consumers for the form of a product.
- the present invention has been made in view of these problems, and an object of the present invention is to provide a technique for evaluating the impression of a product by combining design elements related to the product part and supporting the improvement of the impression. Another purpose is to provide a technique for recommending a product based on the impression evaluation of the product.
- the design support device of a certain aspect of the present invention includes an input unit for inputting a plurality of design element indexes quantifying the design for a plurality of parts constituting the shoe, and a plurality of input units. It includes a prediction unit that predicts an impression index that quantifies the impression of the shoe by combining design element indexes, and an output unit that outputs the predicted impression index.
- Another aspect of the present invention is a design support method.
- an input step for inputting a plurality of design element indexes quantifying the design for a plurality of parts constituting the shoe and a plurality of input design element indexes are combined to quantify the impression of the shoe. It includes a prediction step for predicting an impression index and an output step for outputting the predicted impression index.
- Yet another aspect of the present invention is a design support device.
- This device combines an input unit for inputting a plurality of quantified design element indexes including at least a design for a plurality of parts constituting the garment and a plurality of the input design element indexes to give an impression to the garment. It includes a prediction unit that predicts a plurality of quantified impression indexes and an output unit that outputs the predicted plurality of impression indexes.
- Yet another aspect of the present invention is a recommended device.
- This device uses a pair of a plurality of design element indexes quantifying the design for a plurality of parts constituting the shoe and an impression index quantifying the impression of the shoe as teacher data, and the plurality of the design element index and the impression.
- the impression index is used for a plurality of design element indexes of shoes designated by the user based on the learning unit that learns the correlation between the indexes and generates an impression prediction model and the learned impression prediction model. Includes a prediction unit that predicts the above and a recommendation unit that recommends other shoes having an impression index similar to the predicted impression index.
- Yet another aspect of the present invention is also a recommended device.
- This device uses a pair of a plurality of design element indexes quantifying the design for a plurality of parts constituting the garment and an impression index quantifying the impression on the garment as teacher data, and the plurality of the design element indexes and the impression.
- the impression index is used for a plurality of design element indexes of clothing specified by the user based on the learning unit that learns the correlation between the indexes and generates an impression prediction model and the learned impression prediction model. Includes a prediction unit that predicts the above and a recommendation unit that recommends other clothing having an impression index similar to the predicted impression index.
- the present invention it is possible to evaluate the impression of a product by combining design elements related to the product part and support the improvement of the impression.
- the product can be recommended based on the impression evaluation of the product.
- FIG. It is a block diagram of the design support apparatus which concerns on Embodiment 1.
- FIG. It is a figure which shows an example of the design of the whole shoe. It is a figure which shows the sole of the shoe of FIG. It is a figure which shows the upper of the shoe of FIG. It is a figure which shows the texture attached to the upper of a shoe. It is a figure which shows the radius of curvature of the instep of a shoe. It is a figure which shows the toe elevation angle of a shoe. It is a figure which shows the height of a shoe. It is a figure explaining the example of the impression index for each target segment. It is a figure explaining the correlation between a shoe design element index and an impression index for a certain target segment.
- FIG. 17 (a) to 17 (f) are diagrams for explaining the coloring pattern of the main color of the shoes to be evaluated. It is a block diagram of the product recommended apparatus which concerns on Embodiment 2.
- FIG. 17 (a) to 17 (f) are diagrams for explaining the coloring pattern of the main color of the shoes to be evaluated.
- FIG. 1 is a configuration diagram of a design support device 100 according to the present embodiment.
- the figure depicts a block diagram focusing on functions, and these functional blocks can be realized in various forms by hardware, software, or a combination thereof.
- the design support device 100 evaluates the impression index of the entire shoe according to the target user segment from the design element index of the shoe part.
- the design element index input unit 10 inputs a design element index that quantifies the design of the entire shoe or the parts that make up the shoe.
- the design element index is an index relating to the shape, color, or texture of the entire shoe or the parts constituting the shoe.
- FIG. 2 is a diagram showing an example of the design of the entire shoe. The shoe is divided into two parts, the sole and the upper.
- FIG. 3 is a diagram showing the sole of the shoe of FIG. 2
- FIG. 4 is a diagram showing the upper of the shoe of FIG.
- the upper of the shoe of FIG. 3 is given a plain texture, but a texture such as stripes or polka dots may be given.
- FIG. 5 is a diagram showing a plain texture attached to the upper of the shoe.
- FIG. 6 shows the radius of curvature R of the instep of the shoe
- FIG. 7 shows the toe elevation angle ⁇ of the shoe
- FIG. 8 shows the height H of the shoe.
- the radius of curvature R of the instep of the shoe, the toe elevation angle ⁇ of the shoe, and the height H of the shoe are examples of indicators relating to the shape of the entire shoe.
- the shape of the shoe portion there are the shape of the sole in FIG. 3 and the shape of the upper in FIG.
- the shape of the sole includes the thickness of the entire sole and the change in the thickness of the sole from the toe to the heel.
- the shape of the upper includes the thinness of the toe side and the change in the curvature of the heel curve.
- indexes related to the color of the entire shoe in FIG. 2 there are indexes of overall hue average, overall brightness average, and overall saturation average, which are given by the average values of hue, lightness, and saturation of the entire shoe, respectively.
- index related to the color of the entire shoe there is an index of the center of gravity in the anteroposterior or vertical direction of hue, lightness, and saturation, and the position of the center of gravity of the hue, lightness, and saturation in the anteroposterior or vertical direction of the shoe, respectively. Given in.
- the front-back direction of the shoe is the direction from the toe to the heel of the shoe
- the up-down direction of the shoe is the direction from the upper to the sole of the shoe. Similar color indicators can be defined for shoe parts such as the upper in FIG. 3 and the sole in FIG.
- a plain texture can be added to the upper of the shoe.
- Other examples of textures are patterns such as stripes and polka dots.
- the index related to the texture is an index related to the shape and position of the texture of the shoe part.
- the target segment input unit 20 inputs the segment of the target user.
- Users may be segmented by attributes such as gender, age, place of residence, running frequency, and running history.
- the user base is classified into multiple target segments such as serious runners who are serious about running and fan runners who enjoy running as a hobby.
- the impression index input unit 30 inputs an impression index that quantifies the impression that the user has on the shoes for each target segment.
- impression indicators include sturdiness, speed, and luxury, and each target segment user has an evaluation value of, for example, A (excellent), B (average), and C (inferior). Use the data that evaluated the impression index. Even for shoes with the same design, the evaluation value of the impression index differs depending on the user segment.
- FIG. 9 is a diagram illustrating an example of an impression index for each target segment.
- there are six impression indexes I1 to I6 for the target shoe and there is a difference in the evaluation values of the impression indexes I1 to I6 between the serious runner and the fan runner for the same shoe design.
- the learning unit 40 includes a design element index of the shoe to be learned given by the design element index input unit 10, a target segment given by the target segment input unit 20, and the target given by the impression index input unit 30.
- the impression index of the shoes to be learned by the user of the segment as teacher data
- the correlation between the design element index and the impression index is learned for each target segment to generate an impression prediction model.
- the learning unit 40 uses the pair of the design element index and the impression index of the shoe to be learned as teacher data without distinguishing the user group, and determines the design element index and the impression index. Learn the correlation between them to generate an impression prediction model.
- FIG. 10 is a diagram for explaining the correlation between m design element indexes D1 to Dm of shoes and n impression indexes I1 to In for a serious runner.
- the design element indicators are D1: shoe height, D2: toe elevation angle, D3: overall brightness average, D4: overall saturation average, and the like.
- the impression indexes are I1: B, I2: B, etc. for the combination of the design element indexes D1 to Dm of the target shoes in the first row.
- the impression index is for the combination of I1: A, I2: A, and the design element indexes D1 to Dm of the target shoes in the third row.
- the indexes are I1: A and I2: C, and the impression indexes are I1: B and I2: C for the combination of the design element indexes D1 to Dm of the target shoes in the fourth row.
- FIG. 11 is a diagram for explaining the correlation between m design element indexes D1 to Dm of shoes and n impression indexes I1 to In for a fan runner.
- the values of the design element indexes D1 to Dm of the target shoes in each row are the same as those in FIG. 10, but the impression indexes I1 to In are different from those of the serious runner for the same target shoes.
- Impression indexes are I1: B, I2: C for the target shoes in the first row, I1: A, I2: B for the target shoes in the second row, and impressions for the target shoes in the third row.
- the indexes are I1: B and I2: B, and the impression indexes are I1: C and I2: C for the target shoes in the fourth row.
- a pair of shoe design element indexes D1 to Dm and impression indexes I1 to In for learning for each target segment as shown in FIGS. 10 and 11 is given as teacher data, and the impression indexes I1 are given from the shoe design element indexes D1 to Dm.
- a model that predicts ⁇ In is generated by a decision tree analysis method such as random forest.
- a prediction model based on a decision tree with the design element index as the explanatory variable and the impression index as the objective variable is generated for each impression index.
- a method other than decision tree analysis may be used.
- regression models, neural networks, Bayesian inference, etc. may be used.
- the learning unit 40 stores the impression prediction model generated for each target segment in the impression prediction model storage unit 50.
- the impression index prediction unit 60 receives the target segment from the target segment input unit 20, and reads the impression prediction model of the target segment from the impression prediction model storage unit 50. Based on the impression prediction model of the target segment, the impression index prediction unit 60 combines the design element indexes of the shoes to be evaluated given by the design element index input unit 10 with respect to the shoes to be evaluated by the user of the target segment. Predict impression indicators. For example, when the target segment is a serious runner for a combination of multiple design element indicators of the shoe to be evaluated, the evaluation value of the impression index such as excellent speed but inferior robustness becomes a trained prediction model. It is predicted based on it.
- the target impression index to be improved may be input to the impression index prediction unit 60, and the impression index prediction unit 60 may be configured to output an evaluation value for the target impression index.
- the impression index prediction unit 60 may extract at least one design element index in descending order of influence on the evaluated impression index.
- the design element indexes can be extracted in descending order of the degree of influence on the impression index to be evaluated by referring to the degree of influence of the explanatory variables of the decision tree.
- the output unit 70 outputs the evaluation value of the impression index predicted by the impression index prediction unit 60.
- the evaluation value of the impression index is shown, for example, in three stages of A, B, and C, where A is excellent, B is average, and C is inferior.
- the output unit 70 may output the design element indexes extracted for each impression index in descending order of influence together with the evaluation value of the impression index.
- the output unit 70 may output the design element indexes extracted in descending order of influence for each impression index together with the direction of correction or the range of correction of the design element indexes for improving the impression index. Good.
- FIG. 12 is a diagram showing the predicted evaluation value of the impression index I1 for the shoes to be evaluated and the direction and range of modification of the design element index that affects the impression index I1.
- Impression index I1 is a sense of robustness, and its evaluation value is A.
- Three design element indexes D6: sole lightness average, D7: sole lightness anteroposterior center of gravity, and D8: upper lightness anterior-posterior center of gravity are listed in descending order of influence on the impression index I1.
- the average value of each design element index for the entire shoe is aligned on the vertical axis 140, and the value of each design element index is displayed on the horizontal axis.
- the input value 120a of the design element index D6 of the shoe to be evaluated is 0.4.
- the range of the design element index D6 (referred to as “A zone”) that the shoe having the evaluation value of the impression index I1 can take is indicated by reference numeral 110a, and the intermediate value thereof is indicated by a dotted line.
- the range of the design element index D6 (referred to as “C zone”) that can be taken by the shoe whose evaluation value of the impression index I1 is C is indicated by reference numeral 130a, and the intermediate value thereof is indicated by a dotted line.
- the input value 120a of the design element index D6 of the shoe to be evaluated is a value lower than the average value of the entire shoe, is out of the A zone 110a, and is in the C zone 130a.
- the correction direction of the design element index D6 of the shoe to be evaluated for improving the impression index I1 is indicated by the arrow of reference numeral 150a.
- the design element index D6 of the shoe is changed in the direction of increasing the sole brightness average, that is, in the direction of making the sole brighter as a whole, as a design modification direction.
- the range of correction is such that the average value of the entire shoe is exceeded in the correction direction and the median value of the A zone is exceeded in the correction direction.
- the input value 120b of the design element index D7 is a value higher than the average value of the entire shoe, is out of the A zone 110b, and is in the C zone 130b.
- the correction direction of the design element index D7 for improving the impression index I1 is indicated by the arrow of reference numeral 150b.
- the design element index D7 is shown to change the brightness in the direction of changing the center of gravity in the anteroposterior direction of the sole to the heel side, that is, in the direction of brightening the heel side of the sole as the design correction direction.
- the correction direction is indicated by the arrow of reference numeral 150c.
- the design element index D9 is shown to change the brightness in the direction of changing the center of gravity in the front-back direction of the upper brightness to the toe side, that is, in the direction of brightening the toe side of the upper as a design correction direction.
- FIG. 13 is a diagram showing the predicted evaluation value of the impression index I5 for the shoe to be evaluated and the direction and range of modification of the design element index affecting the impression index I5.
- the impression index I5 is a sense of speed, and its evaluation value is B.
- Two design element indexes D8: upper anterior-posterior center of gravity and D10: upper saturation vertical center of gravity are listed in descending order of influence on the impression index I5.
- the input values 120d, A zone 110d, C zone 130d, and correction direction 150d of the design element index D8 of the shoe to be evaluated are shown.
- the design element index D8 is shown to change the brightness in the direction of changing the center of gravity in the front-back direction of the upper brightness to the toe side, that is, in the direction of brightening the toe side of the upper as a design modification direction.
- the design element index D10 is shown as a design modification direction in which the saturation is changed in the direction in which the center of gravity in the vertical direction of the upper saturation is changed upward, that is, in the direction in which the upper side of the upper is vivid.
- the design element index correction unit 80 corrects the design element index according to the direction of correction presented by the output unit 70.
- the design element index may be modified automatically by the design support device 100, or the operator of the design support device 100 may input a change value of the design element index.
- the design element index correction unit 80 gives the modified design element index to the design element index input unit 10.
- the impression index prediction unit 60 predicts the impression index again based on the prediction model based on the modified design element index, and the output unit 70 outputs a new evaluation value of the impression index. By repeating this modification work, the impression index can be improved and the design can be modified while changing the design element index.
- FIG. 14 is a flowchart showing a procedure for generating a prediction model for predicting an impression index for shoes by machine learning by combining a plurality of design element indexes of shoes. This is a machine learning phase in which the learning unit 40 generates a prediction model using evaluation data for shoes of a large number of users.
- the design element index input unit 10 inputs the design element index of the shoe to be learned (S10).
- the image of the shoe to be learned or the three-dimensional model may be preprocessed to extract the shoe design element index.
- the target segment input unit 20 inputs the target segment for evaluating the shoes to be learned, and the impression index input unit 30 inputs the evaluation value of the impression index by each target segment for the shoes to be learned (S20).
- the learning unit 40 machine-learns the correlation between the design element index of the shoe to be learned and the impression index by each target segment, and generates a prediction model that predicts the impression index from the design element index and the target segment (S30).
- the learning unit 40 stores the generated prediction model in the impression prediction model storage unit 50 for each target segment (S40).
- FIG. 15 is a flowchart showing a procedure for predicting an impression index from a design element index and a target segment using a trained prediction model. This is a phase in which the impression index prediction unit 60 predicts the impression index from the design element index and the target segment using the prediction model generated by the learning unit 40.
- the design element index input unit 10 inputs the design element index of the shoe to be evaluated (S50).
- the target segment input unit 20 inputs a specific target segment (S60).
- the impression index input unit 30 designates an impression index as a target for improvement (S70).
- the impression index prediction unit 60 predicts the impression index of the shoe from the input design element index of the shoe to be evaluated based on the prediction model of the specific target segment read from the impression prediction model storage unit 50 (S80). ..
- the output unit 70 outputs the evaluation value of the impression index targeted for improvement together with the range and direction of modification of the design element index that affects the impression index (S90).
- the design element index correction unit 80 corrects the design element index that affects the impression index according to the correction direction, and the impression index prediction unit 60 re-evaluates the impression index from the corrected design element index based on the prediction model ( S100).
- FIG. 16 is a flowchart illustrating an example of the design modification and re-evaluation process S100 of FIG.
- a procedure for changing the color based on the evaluation result of the impression index will be described.
- the coloring pattern of the main color is selected based on the evaluation result of the impression index (S120).
- 17 (a) to 17 (f) are diagrams for explaining the coloring pattern of the main color of the shoes to be evaluated.
- the shaded area is colored with the main color. Areas that can be colored with the primary color are specified in advance. An area that can be colored with the main color may be extracted by specifying an area that is painted with the same color as the main color based on the different painting patterns of shoes sold in the past. Different coloring patterns as shown in FIGS. 17 (a) to 17 (f) can be obtained depending on whether or not the region that can be colored with the main color is colored with the main color.
- One coloring pattern is selected from the coloring patterns shown in FIGS. 17 (a) to 17 (f) based on the contribution of the design element index output together with the evaluation result of the impression index. For example, when the design element index that contributes to the impression index is the center of gravity in the anteroposterior direction of the upper brightness, and the correction direction of the design element index is the direction of brightening the toes, a pattern in which the toes of the upper are colored with the main color is shown in FIG. 17 ( Select from the coloring patterns a) to (f). If there are multiple design element indicators that contribute to the improvement of the impression index, the one with the highest contribution is given priority. If it is not clear which design element index contributes to the improvement of the impression index, a coloring pattern may be randomly selected.
- a plurality of design patterns are generated by coloring a coloring area other than the main color of the selected coloring pattern with all the color combinations selected from the color palette (S130).
- the impression index prediction unit 60 predicts the impression index from the modified design element index of each design pattern based on the prediction model (S140).
- the output unit 70 outputs the evaluation result of the impression index of each design pattern (S150). This makes it possible to re-evaluate the impression of the modified design pattern.
- shoes are taken as an example of products, but the design support device 100 of the present embodiment is applied to products other than shoes, for example, clothes, and the clothes are divided into parts such as collars and sleeves.
- the impression index for the entire garment may be predicted from the design element index of the garment part.
- the evaluation data regarding the impression that the consumer has on the shoe is used as the teacher data, and the design index of the shoe part is used for each target segment.
- the design support device 100 predicts the impression index of the shoe according to a specific target segment from the design index of the shoe part based on the learned prediction model, and modifies the design element index for improving the impression index.
- the product recommendation device of the second embodiment recommends to the user another product that is similar to the impression of the product that the user has shown interest in or purchased in the past on an online shopping site or the like.
- the product may be any product other than the shoes described in the first embodiment, as long as the impression of the product changes depending on the design of the elements constituting the product, such as clothes and stationery.
- FIG. 18 is a configuration diagram of the product recommended device 200 according to the second embodiment. The configuration and operation common to the design support device 100 of the first embodiment will be omitted as appropriate.
- the design element index input unit 10 inputs a design element index that quantifies the design for the entire product or a part constituting the product.
- the target segment input unit 20 inputs the segment of the target user.
- the impression index input unit 30 inputs an impression index that quantifies the impression of the product held by the user for each target segment.
- the learning unit 40 is a product design element index given by the design element index input unit 10, a target segment given by the target segment input unit 20, and an impression index of the user's product of the target segment given by the impression index input unit 30. Is used as teacher data to learn the correlation between the design element index and the impression index for each target segment, generate an impression prediction model, and store it in the impression prediction model storage unit 50. Segmentation of the product user group is optional, and the learning unit 40 uses the pair of the product design element index and the shoe impression index as teacher data without distinguishing the user group, and uses the design element index and the impression index.
- the impression prediction model may be generated by learning the correlation between the two.
- the product selection unit 94 extracts the design element index of the product specified by the user on the online shopping site or the like from the product database 92 and supplies it to the design element index input unit 10.
- the design element index input unit 10 gives the impression index prediction unit 60 the design element index of the product specified by the user.
- the target segment input unit 20 gives the impression index prediction unit 60 the segment of the user who has specified the product. However, if the user segment of the product is not segmented, it is not necessary to input the target segment.
- the impression index prediction unit 60 receives the target segment from the target segment input unit 20, and reads the impression prediction model of the target segment from the impression prediction model storage unit 50.
- the impression index prediction unit 60 predicts the impression index of the user of the target segment for the product by combining the design element indexes of the products given from the design element index input unit 10 based on the impression prediction model of the target segment.
- the impression index of the user for the product is predicted based on the impression prediction model that does not depend on the target segment.
- the recommendation unit 90 selects another product having an impression index similar to the impression index predicted by the impression index prediction unit 60 from the product database 92 and recommends it to the user.
- the similar range of the impression index can be determined by a predetermined threshold value for each impression index.
- the recommendation unit 90 allows the user to browse other products that are similar in impression index to the specified product but have different designs.
- the design element index of the product specified by the user on the online shopping site or the like is given to the design element index input unit 10, but in the past, the user's past product purchase history and censored product list history are referred to.
- the design element index of the purchased or viewed product may be given to the design element index input unit 10.
- the impression index prediction unit 60 predicts the impression index from the combination of the design element indexes of the products purchased or viewed in the past
- the recommendation unit 90 predicts the impression index similar to the impression index of the product purchased or viewed in the past. Recommend other products with.
- the present invention can be used in a technique for supporting the design of a product and a technique for recommending the product.
- design element index input unit 20 target segment input unit, 30 impression index input unit, 40 learning unit, 50 impression prediction model storage unit, 60 impression index prediction unit, 70 output unit, 80 design element index correction unit, 90 recommendation unit , 92 product database, 94 product selection section, 100 design support device, 200 product recommendation device.
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Business, Economics & Management (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Strategic Management (AREA)
- Evolutionary Computation (AREA)
- Software Systems (AREA)
- Economics (AREA)
- Development Economics (AREA)
- General Engineering & Computer Science (AREA)
- Medical Informatics (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Human Resources & Organizations (AREA)
- General Business, Economics & Management (AREA)
- Game Theory and Decision Science (AREA)
- Entrepreneurship & Innovation (AREA)
- Accounting & Taxation (AREA)
- Finance (AREA)
- Marketing (AREA)
- Mathematical Physics (AREA)
- Data Mining & Analysis (AREA)
- Computing Systems (AREA)
- Computer Hardware Design (AREA)
- Geometry (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Tourism & Hospitality (AREA)
- Footwear And Its Accessory, Manufacturing Method And Apparatuses (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
Description
図1は、本実施の形態に係るデザイン支援装置100の構成図である。同図は機能に着目したブロック図を描いており、これらの機能ブロックはハードウェア、ソフトウェア、またはそれらの組合せによっていろいろな形で実現することができる。
実施の形態2として、実施の形態1で説明した印象予測モデルを利用した商品推奨装置について説明する。実施の形態2の商品推奨装置は、オンラインショッピングサイトなどでユーザが興味を示したり、過去に購入した商品の印象と類似する別の商品をユーザに推奨する。商品は、実施の形態1で説明した靴以外に衣服、文房具など、商品を構成する要素のデザインによって商品の印象が変わる商品であればどんな商品であってもよい。
Claims (14)
- 靴を構成する複数の部分に対するデザインを定量化した複数のデザイン要素指標を入力する入力部と、
入力された複数の前記デザイン要素指標を組み合わせて、前記靴に対する印象を定量化した印象指標を予測する予測部と、
予測された前記印象指標を出力する出力部とを含むことを特徴とするデザイン支援装置。 - 複数の前記デザイン要素指標と前記印象指標の対を教師データとして用いて複数の前記デザイン要素指標と前記印象指標の間の相関関係を学習して印象予測モデルを生成する学習部をさらに含み、
前記予測部は、学習された前記印象予測モデルにもとづいて、入力された複数の前記デザイン要素指標に対して、前記印象指標を予測することを特徴とする請求項1に記載のデザイン支援装置。 - 前記学習部は、複数のセグメントに分類された前記靴のユーザの前記セグメント毎の複数の前記デザイン要素指標と前記印象指標の対を教師データとして用いて複数の前記デザイン要素指標と前記印象指標の間の相関関係を学習し、前記セグメント毎の印象予測モデルを生成し、
前記入力部は、さらに前記セグメントを入力し、
前記予測部は、学習された前記セグメント毎の前記印象予測モデルにもとづいて、入力された前記セグメントおよび複数の前記デザイン要素指標に対して、前記印象指標を予測することを特徴とする請求項2に記載のデザイン支援装置。 - 前記予測部は、前記印象指標に与える影響度合いの高い順に少なくとも一つのデザイン要素指標を抽出し、
前記出力部は、前記印象指標毎に抽出された前記デザイン要素指標を前記印象指標とともに出力することを特徴とする請求項1から3のいずれかに記載のデザイン支援装置。 - 前記出力部は、前記印象指標毎に抽出された前記デザイン要素指標について、前記印象指標を改善するための前記デザイン要素指標の修正の方向を合わせて出力することを特徴とする請求項4に記載のデザイン支援装置。
- 前記出力部は、前記印象指標毎に抽出された前記デザイン要素指標について、前記印象指標を改善するための前記デザイン要素指標の修正の範囲を合わせて出力することを特徴とする請求項4に記載のデザイン支援装置。
- 前記印象指標毎に抽出された前記デザイン要素指標を修正するデザイン要素指標修正部をさらに含み、
前記予測部は、修正された前記デザイン要素指標を用いて前記印象指標を再予測することを特徴とする請求項4に記載のデザイン支援装置。 - 前記デザイン要素指標は、前記靴を構成する複数の部分の各々に対する形状に関する指標を含むことを特徴とする請求項1から7のいずれかに記載のデザイン支援装置。
- 前記デザイン要素指標は、前記靴を構成する複数の部分の各々に対する色彩またはテクスチャに関する指標をさらに含むことを特徴とする請求項8に記載のデザイン支援装置。
- 靴を構成する複数の部分に対するデザインを定量化した複数のデザイン要素指標を入力する入力ステップと、
入力された複数の前記デザイン要素指標を組み合わせて、前記靴に対する印象を定量化した印象指標を予測する予測ステップと、
予測された前記印象指標を出力する出力ステップとを含むことを特徴とするデザイン支援方法。 - 衣服を構成する複数の部分に対するデザインを少なくとも含む定量化された複数のデザイン要素指標を入力する入力部と、
入力された複数の前記デザイン要素指標を組み合わせて、前記衣服に対する印象を定量化した複数の印象指標を予測する予測部と、
予測された前記複数の印象指標を出力する出力部とを含むことを特徴とするデザイン支援装置。 - 複数の前記デザイン要素指標と前記印象指標の対を教師データとして用いて複数の前記デザイン要素指標と前記印象指標の間の相関関係を学習して印象予測モデルを生成する学習部をさらに含み、
前記予測部は、学習された前記印象予測モデルにもとづいて、入力された複数の前記デザイン要素指標に対して、前記印象指標を予測することを特徴とする請求項11に記載のデザイン支援装置。 - 予測された前記印象指標と類似する印象指標を有する他の靴を推奨する推奨部をさらに含むことを特徴とする請求項1または2に記載のデザイン支援装置。
- 予測された前記印象指標と類似する印象指標を有する他の衣服を推奨する推奨部をさらに含むことを特徴とする請求項11または12に記載のデザイン支援装置。
Priority Applications (5)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP2020556821A JP7395499B2 (ja) | 2019-07-17 | 2019-07-17 | デザイン支援装置及びデザイン支援方法 |
| PCT/JP2019/028148 WO2021009880A1 (ja) | 2019-07-17 | 2019-07-17 | デザイン支援装置及びデザイン支援方法 |
| CN201980034919.3A CN112534457B (zh) | 2019-07-17 | 2019-07-17 | 设计支援装置以及设计支援方法 |
| US17/255,969 US12373619B2 (en) | 2019-07-17 | 2019-07-17 | Shoe designing support device that estimates impression index, and shoe designing support method that estimates impression index |
| EP19933213.1A EP3798960A4 (en) | 2019-07-17 | 2019-07-17 | DESIGN AID DEVICE AND DESIGN AID METHOD |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| PCT/JP2019/028148 WO2021009880A1 (ja) | 2019-07-17 | 2019-07-17 | デザイン支援装置及びデザイン支援方法 |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| WO2021009880A1 true WO2021009880A1 (ja) | 2021-01-21 |
Family
ID=74209734
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| PCT/JP2019/028148 Ceased WO2021009880A1 (ja) | 2019-07-17 | 2019-07-17 | デザイン支援装置及びデザイン支援方法 |
Country Status (5)
| Country | Link |
|---|---|
| US (1) | US12373619B2 (ja) |
| EP (1) | EP3798960A4 (ja) |
| JP (1) | JP7395499B2 (ja) |
| CN (1) | CN112534457B (ja) |
| WO (1) | WO2021009880A1 (ja) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2022208718A1 (ja) * | 2021-03-31 | 2022-10-06 | 日本電気株式会社 | 商品デザイン生成支援装置、商品デザイン生成支援方法およびプログラム記憶媒体 |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP3980253B1 (en) | 2019-06-06 | 2024-09-25 | Bluebeam, Inc. | Methods and systems for establishing a linkage between a three-dimensional electronic design file and a two-dimensional design document |
| CN115081043B (zh) * | 2022-06-27 | 2024-03-22 | 广东时谛智能科技有限公司 | 鞋体设计中特征数据的确定方法、装置、设备及存储介质 |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH10124476A (ja) * | 1996-10-23 | 1998-05-15 | Nri & Ncc Co Ltd | 階層的予測モデル構築装置及びその方法 |
| JPH10301981A (ja) * | 1997-04-28 | 1998-11-13 | Tokyo Electric Power Co Inc:The | 建物外観の色彩決定支援システム |
| JP2000235588A (ja) | 1999-02-12 | 2000-08-29 | Matsushita Electric Works Ltd | デザイン構築支援システム |
| JP2000285151A (ja) * | 1999-03-31 | 2000-10-13 | Toyota Motor Corp | デザイン評価装置及びデザイン作成装置並びに方法 |
| WO2011103272A2 (en) * | 2010-02-22 | 2011-08-25 | Nike International Ltd. | Augmented reality design system |
| JP2012155422A (ja) * | 2011-01-24 | 2012-08-16 | Chiba Univ | デザイン支援プログラム |
| JP2012178130A (ja) * | 2011-02-25 | 2012-09-13 | Fourdigit Inc | デザイン評価、デザイン最適化を支援するシステム、及びプログラム |
| WO2017098760A1 (ja) | 2015-12-08 | 2017-06-15 | ソニー株式会社 | 情報処理装置、情報処理方法、およびプログラム |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CA2428079A1 (en) | 2000-11-10 | 2002-07-25 | Affinnova, Inc. | Method and apparatus for dynamic, real-time market segmentation |
| JP4894709B2 (ja) * | 2007-10-04 | 2012-03-14 | 株式会社Ihi | 製品設計支援システム及びコンピュータにおける製品設計支援のための動作方法 |
| CN104424230B (zh) * | 2013-08-26 | 2019-10-29 | 阿里巴巴集团控股有限公司 | 一种网络商品推荐方法及装置 |
| JP6199685B2 (ja) * | 2013-10-03 | 2017-09-20 | Necソリューションイノベータ株式会社 | ファッションコーディネート支援装置及びファッションコーディネート支援システム |
| US11858651B2 (en) * | 2018-10-25 | 2024-01-02 | The Boeing Company | Machine learning model development with interactive feature construction and selection |
-
2019
- 2019-07-17 CN CN201980034919.3A patent/CN112534457B/zh active Active
- 2019-07-17 EP EP19933213.1A patent/EP3798960A4/en not_active Withdrawn
- 2019-07-17 JP JP2020556821A patent/JP7395499B2/ja active Active
- 2019-07-17 US US17/255,969 patent/US12373619B2/en active Active
- 2019-07-17 WO PCT/JP2019/028148 patent/WO2021009880A1/ja not_active Ceased
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH10124476A (ja) * | 1996-10-23 | 1998-05-15 | Nri & Ncc Co Ltd | 階層的予測モデル構築装置及びその方法 |
| JPH10301981A (ja) * | 1997-04-28 | 1998-11-13 | Tokyo Electric Power Co Inc:The | 建物外観の色彩決定支援システム |
| JP2000235588A (ja) | 1999-02-12 | 2000-08-29 | Matsushita Electric Works Ltd | デザイン構築支援システム |
| JP2000285151A (ja) * | 1999-03-31 | 2000-10-13 | Toyota Motor Corp | デザイン評価装置及びデザイン作成装置並びに方法 |
| WO2011103272A2 (en) * | 2010-02-22 | 2011-08-25 | Nike International Ltd. | Augmented reality design system |
| JP2012155422A (ja) * | 2011-01-24 | 2012-08-16 | Chiba Univ | デザイン支援プログラム |
| JP2012178130A (ja) * | 2011-02-25 | 2012-09-13 | Fourdigit Inc | デザイン評価、デザイン最適化を支援するシステム、及びプログラム |
| WO2017098760A1 (ja) | 2015-12-08 | 2017-06-15 | ソニー株式会社 | 情報処理装置、情報処理方法、およびプログラム |
Non-Patent Citations (3)
| Title |
|---|
| NOZAKI NAOYUKI , EIICHI KONNO ,MITSURU SATO , MAKOTO SAKAIRI , TOSHIYUKI SHIBUYA , YUUJI KANAZAWA , SERBAN GEORGESCU: "Application of Artificial Intelligence Technology in Product Design", FUJITSU, vol. 53, no. 4, July 2017 (2017-07-01), pages 43 - 51, XP055784113 * |
| OTA KENICHIRO: "Doesthe design evolve by Using Al?", NIKKEI DESIGN , vol. 378, 24 November 2018 (2018-11-24), pages 52 - 53, XP009526204, ISSN: 0913-3429 * |
| See also references of EP3798960A4 |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2022208718A1 (ja) * | 2021-03-31 | 2022-10-06 | 日本電気株式会社 | 商品デザイン生成支援装置、商品デザイン生成支援方法およびプログラム記憶媒体 |
| JP7485203B2 (ja) | 2021-03-31 | 2024-05-16 | 日本電気株式会社 | 商品デザイン生成支援装置、商品デザイン生成支援方法およびコンピュータプログラム |
Also Published As
| Publication number | Publication date |
|---|---|
| EP3798960A4 (en) | 2022-03-23 |
| CN112534457B (zh) | 2024-10-01 |
| JP7395499B2 (ja) | 2023-12-11 |
| JPWO2021009880A1 (ja) | 2021-01-21 |
| US12373619B2 (en) | 2025-07-29 |
| CN112534457A (zh) | 2021-03-19 |
| US20210173984A1 (en) | 2021-06-10 |
| EP3798960A1 (en) | 2021-03-31 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| EP4184414A1 (en) | Product recommendation device and method based on image database analysis | |
| JP7395499B2 (ja) | デザイン支援装置及びデザイン支援方法 | |
| EP3762795B1 (en) | Method, device, system and program for setting lighting condition and storage medium | |
| KR20190132562A (ko) | 신발 디자인 도구 | |
| Ma et al. | The parameter basin and complex of dynamic game with estimation and two-stage consideration | |
| CN108960959A (zh) | 基于神经网络的多模态互补服装搭配方法、系统及介质 | |
| KR102176786B1 (ko) | 온라인 스타일링 시스템 및 온라인 스타일링 제안 방법 | |
| EP4353116B1 (en) | Shoe fit evaluation device and shoe fit evaluation method | |
| Lee et al. | Soccer shoe recommendation system based on multitechnology integration for digital transformation | |
| Neumann et al. | Evolutionary image transition using random walks | |
| KR20190114448A (ko) | 사용자 맞춤형 의류 추천 방법 및 이를 실행하는 서버 | |
| Zhang et al. | Shoe-last design exploration and customization | |
| Wu et al. | A computer-aided coloring method for virtual agents based on personality impression, color harmony, and designer preference | |
| CN113408798B (zh) | 一种面向色觉异常人群的无障碍vr教学资源色彩优化方法 | |
| Benghanem et al. | Alpine skis categorization with comprehensive decision tree models | |
| JP2017167759A (ja) | コーディネート推薦装置、及びプログラム | |
| Shahrabi et al. | Developing a hybrid intelligent model for constructing a size recommendation expert system in textile industries | |
| Podlasov et al. | Japanese street fashion for young people: A multimodal digital humanities approach for identifying sociocultural patterns and trends | |
| Valentine et al. | Rolling up the leaf node’to new levels of analysis: How algorithmic decision-making changes roles, hierarchies, and org charts | |
| KR20230155306A (ko) | 의류 스타일링 추천 장치 및 추천 방법 | |
| Ranjan et al. | Fashion forward with ai creations using gan | |
| Mulet et al. | Heuristic guidelines to support conceptual design | |
| JP7779977B1 (ja) | 情報処理装置、情報処理方法及び情報処理プログラム | |
| KR102564000B1 (ko) | 수요-공급 속성 기반 잡화 커스텀 보수 추천 시스템 및 방법 | |
| Kobayashi et al. | Reuse of Kansei Evaluation Results for the Aesthetic Design of Different Types of Products. |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| ENP | Entry into the national phase |
Ref document number: 2020556821 Country of ref document: JP Kind code of ref document: A |
|
| ENP | Entry into the national phase |
Ref document number: 2019933213 Country of ref document: EP Effective date: 20201222 |
|
| 121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 19933213 Country of ref document: EP Kind code of ref document: A1 |
|
| NENP | Non-entry into the national phase |
Ref country code: DE |
|
| WWG | Wipo information: grant in national office |
Ref document number: 17255969 Country of ref document: US |