Accessory suggestion device and accessory suggestion method

Abstract

To suggest combination according to a purpose of an outing from among accessories owned by a user.SOLUTION: An accessory suggestion device of the present invention comprises: an input processing unit which accepts conditions regarding accessories worn by a user; an accessory suggestion unit which determines candidates of combination of accessories to be worn by the user from wearing history information storing combination actually worn in the past by the user among combination of the accessories owned by the user on the basis of the accepted conditions; and an output processing unit which displays the determined candidates to an output device.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to a clothing and accessories suggestion device and a clothing and accessories suggestion method. [Background technology]

[0002] Food, clothing, and shelter are the basic necessities of life, and people choose their clothing before going out. Choosing clothing can be both fun and tedious for some people. For those who find it tedious, it is a hassle to remember what clothing they own in the short time before going out, and it is also a hassle to think about how to combine multiple clothing items (coordinate) depending on the purpose of their outing. Furthermore, people tend to forget what coordination they have chosen in the past for what purpose. Although there are experts who can suggest coordination ideas, it is unrealistic to consult with an expert every day.

[0003] The information processing system in Patent Document 1 extracts people who resemble the user from images taken at a location specified by the user and suggests the clothes of those people to the user. The recommendation support system in Patent Document 2 determines products to be suggested to the user from among candidate products sold by businesses based on items (clothing accessories) currently owned by the user and the user's preferences. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Japanese Patent Application Publication No. 2023-95908 [Patent Document 2] Patent Publication No. 2021-140312 Summary of the Invention [Problem to be solved by the invention]

[0005] The information processing system of Patent Document 1 suggests outfits extracted from an image to the user, regardless of what outfits the user currently owns. The recommendation support system of Patent Document 2 goes a step further, encouraging the user to purchase clothing items that the user does not currently own. In either case, a new method was needed to allow users to enjoy the experience of thinking about how to combine outfits they already own. Therefore, an object of the present invention is to propose a combination of clothing accessories owned by a user according to the purpose of going out, etc. [Means for solving the problem]

[0006] The clothing accessory suggestion device of the present invention is characterized by comprising: an input processing unit that receives conditions related to clothing accessories to be worn by a user; a clothing accessory suggestion unit that determines, based on the received conditions, candidate clothing accessory combinations to be worn by the user from wearing history information that stores combinations of clothing accessories owned by the user that the user has actually worn in the past; and an output processing unit that displays the determined candidates on an output device. Other means will be described in the description of the embodiment of the invention. [Effects of the Invention]

[0007] According to the present invention, it is possible to suggest a combination of clothing accessories owned by a user according to the purpose of going out, etc. [Brief explanation of the drawings]

[0008] [Figure 1] FIG. 1 is a diagram illustrating a configuration of a clothing and accessories suggestion device. [Figure 2] FIG. 10 is a diagram showing an example of clothing and accessories information. [Figure 3] FIG. 10 is a diagram illustrating an example of wearing history information. [Figure 4] 1A and 1B are diagrams showing examples of clothing and accessory images and coordinated images; [Figure 5] 10 is a flowchart of a clothing and accessories information creation process. [Figure 6]10 is a flowchart of a wearing history information creation process. [Figure 7] 10 is a flowchart of a learning process procedure. [Figure 8] 10 is a flowchart of a proposal processing procedure. [Figure 9] FIG. 10 is a diagram illustrating an example of a proposal screen. [Figure 10] 10 is a flowchart of an external information acquisition process; [Figure 11] FIG. 10 is a diagram illustrating an example of a proposal screen. DETAILED DESCRIPTION OF THE INVENTION

[0009] Hereinafter, an embodiment of the present invention (hereinafter referred to as "the present embodiment") will be described in detail with reference to the drawings, etc. This embodiment is an example of proposing a combination of accessories to be worn by a person (a user of the clothing accessory suggestion device) when going out. Accessories are items worn by a user on various parts of the body, such as the upper body, lower body, head, hands, feet, and neck, at least a part of which is visible to the outside. Therefore, clothing accessories do not include underwear, but are a concept that includes watches, accessories, etc. Note that the following embodiment uses going out as an example, but the present invention is also applicable to selecting accessories to be worn not only when going out but also when having guests over (situations where it is assumed that people will be seen), etc.

[0010] (Clothing product suggestion device) 1 is a diagram showing the configuration of a clothing and accessories suggestion device 1. The clothing and accessories suggestion device 1 is a general computer, and includes a central control unit 11, input devices 12 such as a camera, microphone, mouse, and keyboard, output devices 13 such as a speaker and display, a main memory device 14, an auxiliary memory device 15, and a communication device 16.

[0011] The input processing unit 21, clothing item suggestion unit 22, and output processing unit 23 in the main memory device 14 are programs. In the following description, when the subject is described as "XX unit," it means that the central control unit 11 reads the program from the auxiliary memory device 15 to the main memory device 14 and realizes the program's functions. The auxiliary memory device 15 stores clothing item information 31, wearing history information 32, proposed models 33, clothing item images 34, and coordinated images 35 (described in detail later). The auxiliary memory device 15 may be configured as an independent unit in a separate housing.

[0012] The clothing suggestion device 1 is connected to a user terminal device 2, an external information device 3, and a camera 8 via a network 4. The user terminal device 2 is a portable or stationary computer used by a user, and has a central control device, input devices (camera, touch panel, microphone, etc.), output devices (speaker, display, etc.), a main memory device, an auxiliary memory device, and a communication device (not shown). The external information device 3 is any device other than the clothing suggestion device 1 and the user terminal device 2, and includes servers of government agencies, companies, etc. that store past, present, and future weather information.

[0013] Many clothing items are washed from time to time, during which the user cannot wear the clothing items. Therefore, the clothing item recommendation device 1 needs to recognize the state of the clothing items in real time (described in detail later). To achieve this, the camera 8 is placed, for example, near a storage cabinet (such as a chest of drawers) 5, a washing machine 6, and a clothes drying area 7. The camera 8 captures images of the clothing items present (taking in or out) in that location and transmits the captured images together with its own location information to the clothing item recommendation device 1. The clothing items are uniquely identified by analyzing the captured images (based on color, pattern, material, shape, etc.).

[0014] To identify clothing accessories, instead of or in addition to the camera 8, RFID (Radio Frequency Identification) readers provided in the storage shed 5, washing machine 6, and clothes drying area 7 may read RFID tags attached to the clothing accessories. The RFID tags store clothing IDs. The RFID readers read the clothing IDs of clothing accessories present (to be taken in or out) at the locations, and transmit the read clothing IDs to the clothing accessory recommendation device 1 together with their own location information.

[0015] (Clothing information) 2 is a diagram showing an example of clothing and accessories information 31. In the clothing and accessories information 31, a clothing and accessories ID stored in a clothing and accessories ID field 101 is associated with a major category in a major category field 102, a minor category in a minor category field 103, a shape in a shape field 104, a storage position in a storage position field 105, a color in a color field 106, a pattern in a pattern field 107, a material in a material field 108, a season in a season field 109, a clothing accessory image in an image field 110, and a condition in a condition field 111.

[0016] The clothing and accessories ID in the clothing and accessories ID column 101 is an identifier that uniquely identifies each individual clothing and accessories. The major categories in the major category column 102 are relatively high-ranking categories among the types of clothing and accessories. The small categories in the small category column 103 are relatively low-level categories among the types of clothing and accessories. The large categories and small categories are freely set according to the convenience of the user. There may be more levels of categories. The shape in the shape column 104 is a characteristic of the shape of the clothing or accessory. In addition to or instead of the characteristic of the shape, the user's body size (height, weight, bust size, waist size, hip size) may be stored. The storage location in the storage location column 105 is the location where the user physically stores the clothing item (the name of the storage cabinet 5). For example, in FIG. 2, "chest 1," "chest 2," etc., uniquely identify each of the chests of drawers that serve as storage cabinets 5 when there are multiple chests of drawers. The granularity of the storage location is arbitrary. For example, the storage location may be "the ○th drawer from the top of chest of drawers 1." The other person flag "(other person)" will be described later.

[0017] The color in the color column 106 is the color of the clothing item. The patterns in the pattern column 107 are patterns for clothing and accessories. The material in the material column 108 is the material of the clothing item. The season in the season column 109 is the season in which the clothing item is most frequently worn. The clothing image in the image field 110 is the name and extension of the clothing image 34, which is an image of an individual clothing item. "#" is an abbreviation for a different character. The clothing image (name and extension) here is meta information attached to the clothing image 34 itself, and the clothing image 34 itself is stored in the auxiliary storage device 15. Of course, the clothing information 31 itself may directly store the clothing image 34.

[0018] The status of the status column 111 is one of "storing", "washing", and "-". "storing" indicates that the clothing item is currently stored in a storage cabinet 5 or the like, and therefore the user can wear the clothing item. "washing" indicates that the clothing item is currently stored in a washing machine 6, and therefore the user cannot wear the clothing item. "-" indicates that the user does not own the clothing item. Note that "wearing" may also be one of the statuses (details will be described later).

[0019] (Wearing history information) 3 is a diagram showing an example of the wearing history information 32. In the wearing history information 32, in association with the wearing occasion ID stored in the wearing occasion ID field 121, the date field 122 stores the date, the destination field 123 stores the destination, the purpose field 124 stores the purpose, the companion field 125 stores the companion, the weather field 126 stores the weather, the evaluation field 127 stores the evaluation, and the worn clothing field 128 stores the worn clothing.

[0020] The wearing occasion ID in the wearing occasion ID column 121 is an identifier that uniquely identifies a wearing occasion. A wearing occasion means one outing, and the combination of clothing accessories cannot be changed during one wearing occasion. Among the wearing occasion IDs, those that start with "F" are called "other person's wearing occasion IDs." Other person's wearing occasion IDs will be described later.

[0021] The date in the date column 122 is the date on which the user went out. The clothing and accessories suggestion unit 22 can identify the season based on this date. The destination in the destination field 123 is the destination of the outing. If the user is receiving guests at home, "home" can also be a destination. The purpose in the purpose column 124 is the purpose of going out. The accompanying person in the accompanying person column 125 is a person who accompanies the user when going out. The weather in the weather column 126 is meteorological information for the destination on the date of going out, and in this case, is the weather (sunny, rainy, snowy, . . . ) and temperature.

[0022] The evaluation in the evaluation column 127 is an evaluation by the user or a companion of the combination of clothing and accessories worn. A combination that is suitable for the destination, purpose, companion (and their clothing), weather, etc., and that well reflects the user's own personality, is highly rated. The evaluation may be a ranking such as A>B>C>..., with "A" being the highest evaluation. The evaluation may also be a score. The higher the score, the higher the evaluation. The worn clothing and accessories in the Worn Clothing and Accessories column 128 are the clothing and accessories IDs of the clothing and accessories that the user actually wore on the occasion. Here, one clothing and accessories ID is stored for each major category. "#" is an abbreviation for representing different clothing and accessories IDs. For example, if a hat was not worn when going out, the Hat column will be blank.

[0023] Fig. 4 shows examples of clothing and accessories images 34 and coordinated images 35. The clothing and accessories image 34 on the top left of Fig. 4 is a color image captured from the front by a camera of a "polo shirt" owned by a certain user (male). The clothing and accessories image 34 on the top right of Fig. 4 is a color image captured from the front by a camera of a "blouse" owned by another user (female).

[0024] The coordinated image 35 on the bottom left of Fig. 4 is a color image captured by a camera from a generally frontal angle, showing a user (male) wearing a combination of multiple clothing items, or a color image synthesized by the clothing item suggestion unit 22. The user is wearing a "dress shirt," "jeans," and "sneakers." This combination is appropriate, for example, for a young man going for a walk in the park in summer.

[0025] The coordinated image 35 on the bottom right of FIG. 4 is a color image captured by a camera from approximately the front, showing a user (female) wearing a combination of multiple clothing items, or a color image synthesized by the clothing item suggestion unit 22. The user is wearing a "T-shirt," a "long skirt," and "pumps." This combination is appropriate, for example, for a mother attending a school class in the summer. The camera that captures the clothing item image 34 and the coordinated image 35 may be the camera 8, or may be a camera that serves as the input device 12.

[0026] (Proposed model) Now, imagine a function F that outputs a combination of clothing items when certain conditions are input. Clothing combination = F (condition)

[0027] The function F is the proposed model 33. The input variable "conditions" includes at least one of the date, destination, purpose, accompanying persons, and weather in the wearing history information 32 (Fig. 3). The output variable "combination of clothing accessories" includes at least one of the combinations of worn clothing accessories (clothing accessory IDs) in the wearing history information 32 (Fig. 3). Figs. 2 and 3 are linked using the clothing accessory ID as a key. Therefore, the clothing accessory suggestion unit 22 can synthesize a coordinated image 35 like the one shown in the lower part of Fig. 4 based on the output of the proposed model 33.

[0028] The proposed model 33 may or may not be an artificial intelligence such as a neural network.

[0029] The proposed model 33 as a neural network has an input layer, multiple intermediate layers, and an output layer. Each intermediate layer has multiple nodes. A "propagation parameter" is defined for each node in the intermediate layer, defining which part of the information received by the node is passed to which node in the next intermediate layer. The clothing suggestion unit 22 optimizes the propagation parameter by machine learning the proposed model 33 using supervised learning data. The supervised learning data is a combination of past values of the input variables and past values of the output variables. Ultimately, the wearing history information 32 becomes the supervised learning data.

[0030] There are also cases where the proposed model 33 is not a neural network. In this case, the proposed model 33 is an application that searches the wearing history information 32 using "at least one of the date, destination, purpose, accompanying persons, and weather" specified by the user as a search key, and acquires the most frequently matching combination of clothing and accessory IDs.

[0031] Whether the proposed model 33 is a neural network or an application, the clothing suggestion unit 22 vectorizes (or scalarizes) "at least one of date, destination, purpose, accompanying persons, and weather" specified by the user, and uses the result as a search key. The clothing suggestion unit 22 vectorizes (or scalarizes) "at least one of date, destination, purpose, accompanying persons, and weather" in the wearing history information 32 in advance, and sets it as a search key. The proposed model 33 can identify a record in the wearing history information 32 that has a search key that is most similar to the search key. The proposed model 33 can quantify the degree of similarity using known methods (cosine similarity, angle between vectors, etc.).

[0032] (Flexibility of clothing proposal department and proposal model) Normally, it is impossible for a search key to completely match any of the searched keys. If the proposed model 33 were to stop outputting data, the usability of the clothing and accessories suggestion device 1 would deteriorate. Therefore, the clothing and accessories suggestion unit 22 artificially changes the values of the input variables slightly and then repeatedly attempts to input the values to the proposed model 33.

[0033] Assume now that the input variable is a five-dimensional vector “[♭,♭,♭,♭].” The number of dimensions, “5,” corresponds to the number of dimensions of the input variables (date, destination, purpose, accompanying persons, and weather). “♭” abbreviates different values. The clothing suggestion unit 22 creates multiple “[♭±a,♭±b,♭±c,♭±d,♭±e].” Here, a, b, c, … are correction values for the input variables, and the clothing suggestion unit 22 slightly changes the values of a, b, c, …. In other words, the clothing suggestion unit 22 creates multiple approximations that allow for a slight range of difference from the true values of the input variables. Then, the proposed model 33 can output multiple combinations of clothing IDs in response to a single input. Specifically, the proposed model 33 outputs multiple combinations in which only the shoes, for example, are different, for a certain input.

[0034] The processing procedures of this embodiment are explained below. There are five processing procedures: a clothing and accessories information creation processing procedure (FIG. 5), a wearing history information creation processing procedure (FIG. 6), a learning processing procedure (FIG. 7), a proposal processing procedure (FIG. 8), and an external information acquisition processing procedure (FIG. 10).

[0035] (Clothing information creation process) 5 is a flowchart of the clothing and accessories information creation process, which is executed when a user takes inventory of the clothing and accessories they own, and when they subsequently acquire new clothing and accessories.

[0036] In step S201, the input processing unit 21 of the clothing accessory suggestion device 1 accepts the clothing accessory image 34. Specifically, the input processing unit 21 accepts an image of a certain clothing accessory taken by the user using the input device 12 (camera) of the clothing accessory suggestion device 1, the input device (camera) of the user terminal device 2, or the camera 8.

[0037] In step S202, the input processing unit 21 accepts clothing attributes. Specifically, the input processing unit 21 accepts the clothing attributes of the clothing item input by the user via the input device 12 (keyboard) of the clothing suggestion device 1 or the input device (touch panel) of the user terminal device 2. The clothing attributes are information to be stored in the major category column 102 to the season column 109 of the clothing information 31 (FIG. 2). The input processing unit 21 may automatically acquire some of the clothing attributes (shape, color, pattern, etc.) by analyzing the image, without waiting for an input by the user.

[0038] In step S203, the clothing accessory proposal unit 22 of the clothing accessory proposal device 1 creates a record of the clothing accessory information 31 (FIG. 2). Specifically, the clothing accessory proposal unit 22 creates one record of the clothing accessory information 31 using the information received in steps S201 and S202 and the clothing accessory ID assigned by itself. At this stage, the status column 111 is blank. As the clothing accessory moves, the clothing accessory proposal unit 22 receives information from the camera 8, and stores and updates the status in real time.

[0039] In step S204, the clothing accessory suggestion unit 22 determines whether or not all clothing accessory images 34 have been received. Specifically, if all clothing accessory images 34 have been received (step S204 "YES"), the clothing accessory suggestion unit 22 proceeds to step S205. Otherwise (step S204 "NO"), the process returns to step S201.

[0040] In step S205, the clothing and accessories suggestion unit 22 stores the clothing and accessories information 31. Specifically, the clothing and accessories suggestion unit 22 compiles a plurality of records and stores them in the auxiliary storage device 15 as the clothing and accessories information 31. Thereafter, the clothing and accessories information creation process procedure ends.

[0041] (Worn history information creation process procedure) 6 is a flowchart of the wearing history information creation process, which is executed every time the user goes out.

[0042] In step S301, the input processing unit 21 of the clothing and accessories suggestion device 1 accepts destinations and the like. Specifically, the input processing unit 21 accepts the user's input of destinations and the like via the input device 12 (keyboard, microphone) of the clothing and accessories suggestion device 1 or the input device (touch panel, microphone) of the user terminal device 2. The "destination, etc." is information to be stored in the date column 122 to the weather column 126 of the wearing history information 32 (FIG. 3). If the destination, etc. is stored in the external information device 3 (scheduler, etc.), the input processing unit 21 may acquire the destination, etc. from the external information device 3. If the destination is far from the user's home, the weather at the destination is often unknown. In this case, the user does not need to input the weather. However, it is desirable for the user to input an address or the like that identifies the location as the destination.

[0043] Step S301 is performed immediately before the user decides on a clothing combination and goes out. The user may operate the clothing recommendation device 1 and receive suggestions for clothing combinations from the clothing recommendation device 1, or may decide on a clothing combination based on their own judgment without using the clothing recommendation device 1. In order to reflect changes in the user's own preferences, which do not rely on suggestions from the clothing recommendation device 1, in the supervised learning data, it is preferable to allow the user to make their own decision. After that, when the user actually wears the clothing combination, the process proceeds to step S302.

[0044] In step S302, the input processing unit 21 accepts the coordinated image 35. Specifically, after the user puts on all the clothing accessories, the input processing unit 21 accepts an image of the user's entire body taken by the input device 12 (camera) of the clothing accessories suggestion device 1, the input device (camera) of the user terminal device 2, or the camera 8 as the coordinated image 35.

[0045] In step S303, the clothing suggestion unit 22 of the clothing suggestion device 1 creates a record of the wearing history information 32 (FIG. 3). Specifically, the clothing suggestion unit 22 creates one record of the wearing history information 32 using the information received in steps S301 and S302 and the wearing occasion ID numbered by the clothing suggestion unit 22. At this stage, the evaluation column 127 is blank. The weather column 126 may also be blank. The clothing suggestion unit 22 analyzes the coordination image 35 received in step S302, thereby obtaining the clothing IDs of the multiple clothing items appearing in the coordination image 35 from the clothing information 31. Incidentally, at this stage, the clothing suggestion unit 22 updates the status (FIG. 2) of the clothing items appearing in the coordination image 35 to "being worn."

[0046] In step S304, the clothing and accessories recommendation unit 22 determines whether the weather is clear. Specifically, if the weather column 126 of the record created in step S303 is blank (step S304 "NO"), the clothing and accessories recommendation unit 22 proceeds to step S305. Otherwise (step S304 "YES"), the process proceeds to step S306.

[0047] In step S305, the clothing and accessories suggestion unit 22 acquires weather information from the external information device 3. Specifically, the clothing and accessories suggestion unit 22 searches the external information device 3 using the destination (address, etc.) as a key, and acquires the weather information at the destination.

[0048] In step S306, the clothing accessory suggestion unit 22 accepts the most recent evaluation. Specifically, the clothing accessory suggestion unit 22 accepts an evaluation of the clothing accessory combination in the most recent outing, input via the input device 12 (keyboard) of the clothing accessory suggestion device 1 or the input device (touch panel) of the user terminal device 2.

[0049] In step S307, the clothing accessory suggestion unit 22 stores the record of the wearing history information 32. Specifically, the clothing accessory suggestion unit 22 adds the weather acquired in step S305 to the record created in step S303 as necessary, and accumulates the added information in the wearing history information 32. Furthermore, the clothing accessory suggestion unit 22 stores the evaluation received in step S306 in the evaluation column 127 of the record of the most recent outing occasion. Thereafter, the wearing history information creation process procedure ends.

[0050] (Learning process procedure) 7 is a flowchart of the learning process procedure. The learning process procedure is repeatedly executed at any timing after a predetermined number of records of the wearing history information 32 have been accumulated. In the following description, the proposed model 33 is assumed to be a neural network.

[0051] In step S401, the clothing and accessories suggestion unit 22 of the clothing and accessories suggestion device 1 acquires the untrained proposed model 33. Specifically, the clothing and accessories suggestion unit 22 acquires the untrained proposed model 33 from the auxiliary storage device 15.

[0052] In step S402, the clothing and accessories suggestion unit 22 acquires supervised learning data. Specifically, the clothing and accessories suggestion unit 22 acquires the latest wearing history information 32 from the auxiliary storage device 15. Of the wearing history information 32, the information in the date column 122 to the weather column 126 becomes input-side learning data, and the information in the worn clothing and accessories (clothing accessory ID) column 128 becomes output-side learning data.

[0053] In step S403, the clothing and accessories suggestion unit 22 performs machine learning on the proposed model 33. Specifically, the clothing and accessories suggestion unit 22 optimizes the propagation parameters of the proposed model 33 using the supervised learning data acquired in step S402.

[0054] In step S404, the clothing and accessory proposal unit 22 stores the trained proposed model 33. Specifically, the clothing and accessory proposal unit 22 stores the proposed model 33 with the optimized propagation parameters together with the propagation parameters in the auxiliary storage device 15. Thereafter, the learning process procedure ends.

[0055] (Proposal processing procedure) 8 is a flowchart of the proposed processing procedure. The proposed processing procedure is executed every time the user goes out. In the explanation of FIG. 8, reference will be made to FIG. 9.

[0056] In step S501, the clothing accessory suggestion unit 22 of the clothing accessory suggestion device 1 acquires the trained proposed model 33. Specifically, the clothing accessory suggestion unit 22 acquires the trained proposed model 33 from the auxiliary storage device 15.

[0057] In step S502, the input processing unit 21 of the clothing and accessories suggestion device 1 accepts destinations and the like. Specifically, the input processing unit 21 accepts the user's input of destinations and the like via the input device 12 (keyboard, microphone) of the clothing and accessories suggestion device 1 or the input device (touch panel, microphone) of the user terminal device 2. "Destinations and the like" are information to be stored in the date column 122 to the weather column 126 of the wearing history information 32 (FIG. 3). If the destinations and the like are stored in the external information device 3 (scheduler, etc.), the input processing unit 21 may acquire the destinations and the like from the external information device 3. If the destination is far from the user's home, the weather at the destination is often unknown. In this case, the user does not need to input the weather. However, it is desirable for the user to input an address or the like that identifies the location as the destination.

[0058] The timing and processing content of step S502 are the same as those of step S301 (FIG. 6). The processing performed here corresponds to step S301 from the viewpoint of creating the wearing history information 32, and corresponds to step S502 from the viewpoint of receiving suggestions for clothing and accessory combinations. The two are described separately solely for the sake of convenience.

[0059] In step S503, the clothing and accessories proposal unit 22 inputs the destination, etc. to the proposed model 33. Specifically, the clothing and accessories proposal unit 22 inputs the "destination, etc." to the trained proposed model 33. The destination, etc. are "conditions" input to the proposed model 33 (function F). If the weather is unknown, the clothing and accessories proposal unit 22 acquires the weather from the external information device 3 using the method described above.

[0060] In step S504, the clothing accessory suggestion unit 22 acquires a coordinated image 35 from the proposed model 33. Specifically, first, the clothing accessory suggestion unit 22 receives the output of the worn clothing accessory as a combination of multiple clothing accessory IDs from the trained proposed model 33. As described above, multiple combinations are output here. The combination of clothing accessory IDs output here is stored in the worn clothing accessory (clothing accessory ID) column 128 of any one of the records of the wearing history information 32.

[0061] Second, the clothing and accessories suggestion unit 22 deletes, from the combinations output in the "first" of step S504, any combination in which at least one clothing and accessories included in the combination is in a state other than "stored." Third, the clothing and accessories suggestion unit 22 identifies clothing and accessories images 34 based on the combination of clothing and accessories IDs that remain without being deleted in the “second” step of step S504, and synthesizes the coordinated image 35. Here, a plurality of coordinated images 35 are synthesized.

[0062] In step S505, the clothing and accessories suggestion unit 22 ranks the coordination images 35 based on the evaluation. Specifically, the clothing and accessories suggestion unit 22 ranks the coordination images 35 in descending order of evaluation.

[0063] In step S506, the output processing unit 23 of the clothing and accessories suggestion device 1 displays the coordinate images 35. Specifically, the output processing unit 23 displays the coordinate images 35 in descending order of evaluation on the output device 13 (display) of the clothing and accessories suggestion device 1 or the output device (display) of the user terminal device 2. At this time, the output processing unit 23 displays a suggestion screen 41. Although this is midway through FIG. 8, the description will now move to FIG. 9.

[0064] (Proposal screen) 9 is a diagram showing an example of the proposal screen 41. The explanation field 42 of the proposal screen 41 displays guidance text including the destination and the like input by the user in step S502. The proposal screen 41 displays a predetermined number of coordinated images 35 ranked in step S505 in descending order of evaluation. The predetermined number here is "1."

[0065] Now, let us assume that the user positions the cursor 46 on the dress shirt using the input device 12 (or a finger), such as a mouse, and then presses the "Check storage location" button 43. In response, the output processing unit 23 displays the answer "Dress shirt is stored in chest of drawers 1" near the button. As mentioned above, "chest of drawers 1" identifies one of the multiple chests of drawers that exist as storage cabinets 5.

[0066] Next, the user moves the cursor 46 with the input device 12 (or a finger) such as a mouse to shoes and presses the "View other candidates" button 44. The output processing unit 23 then displays the answer "Shoes (hiking boots), stored in shoebox 1" near the button. The hiking boots are included in the combination of clothing items that was rated next. As mentioned above, "shoebox 1" specifies one of the multiple shoeboxes that exist as storage sheds 5.

[0067] Furthermore, suppose the user presses the "History Inquiry" button 45. In response, the output processing unit 23 displays the answer "Year, Month, Day, B Department Store, Rating: A" near the button. This is the destination, etc., that corresponds to the coordinated image 35 displayed here. Return to FIG. 8 for explanation.

[0068] In step S507, the clothing and accessories suggestion unit 22 determines whether or not the inquiry has been received. Specifically, if the user presses any one of the "storage location inquiry" button 43, the "other candidate inquiry" button 44, and the "history inquiry" button 45 on the suggestion screen 41 displayed in step S506 (step S507 "YES"), the clothing and accessories suggestion unit 22 proceeds to step S508. Otherwise (step S507 "NO"), the suggestion processing procedure is terminated.

[0069] In step S508, the clothing and accessories suggestion unit 22 responds to the inquiry. Specifically, the clothing and accessories suggestion unit 22 displays a reply message as described in Fig. 9. After that, the suggestion process procedure ends.

[0070] (External information acquisition processing procedure) 10 is a flowchart of the external information acquisition process. The external information acquisition process is executed at any timing. Many users are interested in what kind of clothing accessories are currently popular in town, etc.

[0071] In step S601, the clothing accessory suggestion unit 22 of the clothing accessory suggestion device 1 connects to the external information device 3. Specifically, the clothing accessory suggestion unit 22 connects the clothing accessory suggestion device 1 or the user terminal device 2 to the external information device 3 via the network 4 in response to a user's instruction. The user views any site, such as a news article or someone else's SNS (Social Networking Service), via the output device 13 of the clothing accessory suggestion device 1 or the output device of the user terminal device 2. The site displays images of other people going out. Now, let's assume that the user selects a certain person's outfit image 35 on a certain site.

[0072] In step S602, the clothing and accessories suggestion unit 22 identifies an image. Specifically, the clothing and accessories suggestion unit 22 identifies the coordinated image 35 selected by the user and cuts it out from the background.

[0073] In step S603, the clothing and accessories suggestion unit 22 extracts individual clothing and accessories. Specifically, the clothing and accessories suggestion unit 22 breaks down the coordinated image 35 cut out from the background into large categories, for example, and extracts multiple clothing and accessories images 34.

[0074] In step S604, the clothing and accessories suggestion unit 22 identifies the destination, etc. Specifically, first, the clothing and accessories suggestion unit 22 identifies the date, destination, purpose, accompanying persons, and weather related to the coordinated image 35 extracted in step S602 from the metadata, etc. attached to the image. Secondly, if the weather information cannot be acquired, the clothing and accessories suggestion unit 22 acquires weather information for the date and destination from any other site (such as a government agency site).

[0075] In step S605, the clothing suggestion unit 22 creates a record of the wearing history information 32 (FIG. 3). Specifically, the clothing suggestion unit 22 creates a record of the wearing history information 32 using the clothing images 34 extracted in step S603 and the destinations, etc. identified in step S604, and stores the record in the auxiliary storage device 15. The clothing suggestion unit 22 assigns a clothing ID to each clothing image 34, and assigns a wearing occasion ID starting with "F" to each coordination image 35. The worn clothing in the record of the wearing history information 32 stored in this way reflects the user's preferences. Of course, the user does not own the worn clothing here.

[0076] In step S606, the clothing accessory proposal unit 22 creates a record of the clothing accessory information 31 (FIG. 2). Specifically, the clothing accessory proposal unit 22 creates a record of the clothing accessory information 31 for each clothing accessory image 34 and stores it in the auxiliary storage device 15. At this time, the clothing accessory proposal unit 22 analyzes the clothing accessory image 34 to determine the major category, minor category, shape, color, pattern, and season. The clothing accessory proposal unit 22 stores "(other person)" as the storage location. The clothing accessory proposal unit 22 accepts the material input by the user and stores it.

[0077] In step S607, the clothing and accessories suggestion unit 22 calculates the degree of similarity with the owned clothing and accessories. Specifically, first, the clothing and accessories suggestion unit 22 acquires a clothing image 34 from the image column 110 of the record stored in step S606. The acquired clothing image 34 is called a "comparison reference image." Second, the clothing and accessories suggestion unit 22 acquires clothing and accessories images 34 from the image column 110 of all records in the clothing and accessories information 31 that have the same major category as the comparison reference image and do not have an other person flag. The clothing and accessories images 34 acquired here are called "comparison target images." Third, the clothing / accessories suggestion unit 22 compares the comparison reference image with each of the comparison target images using an arbitrary method, and calculates the degree of similarity.

[0078] In step S608, the clothing suggestion unit 22 associates similar clothing items. Specifically, the clothing suggestion unit 22 creates "similarity information" that associates clothing items that are similar enough to meet a predetermined standard, and stores the information in the auxiliary storage device 15. The similarity information is, for example, "[D004, (D###, similarity #), (D###, similarity #),...]". Among these, "D004" is the clothing ID that identifies the clothing item in the comparison reference image. "(D###, similarity #)" is the clothing ID that identifies the clothing item in the comparison target image and its similarity. "#" abbreviates to indicate a different value. The external information acquisition process then ends.

[0079] In this way, once records of clothing accessory information 31 (Figure 2) and wearing history information (Figure 3) are accumulated, when the user goes out, they can view coordination images 35 (Figure 9) on the suggestion screen 41, which include suggestions for clothing accessories that the user owns and clothing accessories that the user does not own.

[0080] In this case, when the user selects an item of clothing that the user does not own with the cursor 46 on the proposal screen 41 and then presses the "Check storage location" button 43, the output processing unit 23 displays "(other person)." The user then knows that the item of clothing is not owned by the user. In this case, if the user further presses the "Check other candidate items" button 44, the output processing unit 23 displays candidate items of clothing that the user owns in descending order of similarity to the selected item of clothing. The user can then identify an outfit that they can wear immediately based on the outfits of others.

[0081] (Variation 1: Multidimensional evaluation) The ratings in the wearing history information 32 may be a multidimensional vector such as "(own rating, other person 1's rating, other person 2's rating, ...), average value." Other person 1, other person 2, ... are people other than the user and are designated by the user in advance. The average value of the last dimension is the average value of own rating, other person 1's rating, other person 2's rating, .... The external information device 3 in FIG. 1 also includes a computer operated by a person other than the user. In step S306 in FIG. 6, the clothing suggestion unit 22 accepts an input of a rating of the clothing combination in the most recent outing by a person other than the user via an input device (such as a touch panel) of the external information device 3. The person other than the user may or may not be a companion in the previous outing.

[0082] In this way, the user can know the following combinations, for example: Combinations that were surprisingly highly (lowly) rated by others Combinations that were highly (low) rated by specific others

[0083] (Variation 2: Other proposal screen) Immediately after step S504 (FIG. 8), the clothing / accessories suggestion unit 22 and the output processing unit 23 may perform the following process. <1> The clothing and accessories suggestion unit 22 identifies the clothing and accessories that appear most frequently from among the multiple coordinated images 35 synthesized in the “third” step S504. This clothing and accessories is called the “favorite clothing and accessories.” <2> The clothing and accessories suggestion unit 22 identifies the major category or minor category that appears most frequently together with the favorite clothing and accessories in the multiple coordination images 35. The major category or minor category is called a “flattering category.” In the multiple coordination images 35, each clothing and accessories belonging to the flattering category is called a “flattering clothing and accessories.” <3> The output processing unit 23 displays the favorite clothing item and a plurality of matching clothing items as a suggestion screen on the output device 13 of the clothing item suggestion device 1 or the output device of the user terminal device 2.

[0084] FIG. 11 is a diagram showing an example of the suggestion screen 41 (41a, 41b). The explanation field 51 of the suggestion screen 41b displays a guide message that guides the user to favorite clothing and prompts the user to select from a plurality of candidates a complementing clothing to be combined with the favorite clothing. The favorite clothing is displayed in field 52 of the suggestion screen 41b. Complementing clothing is displayed in fields 53a, 53b, and 53c.

[0085] A user who sees the proposal screen 41b knows the following. For example, when I went shopping in the winter, I often wore the sweater in column 52. - When wearing the sweater in column 52, I was wearing the hats in columns 53a, 53b and 53c. I will continue to treasure the sweater in column 52, and when I buy a new hat that will make the most of this sweater, I should choose one in the same style as the hats in columns 53a, 53b, and 53c.

[0086] The same information as above is displayed in the explanation field 54 of the suggestion screen 41c. Favorite clothing and accessories are displayed in field 55 of the suggestion screen 41c. Complementary clothing and accessories are displayed in fields 56a, 56b, and 56c.

[0087] A user who sees the proposal screen 41c knows the following. - For example, when going to parties held at hotels, she frequently wore high heels (see column 55). - When wearing the high heels in column 55, she was wearing the dresses in columns 56a, 56b and 56c. I will continue to treasure the high heels in column 55, and when I buy a new dress that will allow me to wear them to the fullest, I should choose one that is in the same style as the dresses in columns 56a, 56b, and 56c.

[0088] (Effects of the embodiment) The clothing and accessories suggestion device of this embodiment has the following advantages. (1) The clothing and accessories suggestion device can suggest a combination of clothing and accessories owned by the user that matches the conditions under which the user goes out. (2) The clothing and accessories suggestion device can use the proposed model (freedom of model selection). (3) The clothing and accessories recommendation device can machine-learn the proposed model using appropriate supervised learning data at appropriate times. (4) The clothing and accessories suggestion device can suggest clothing and accessories combinations using images. (5) The clothing and accessories suggestion device can provide feedback on the evaluation of the clothing and accessories combination.

[0089] The present invention is not limited to the above-described embodiments and includes various modifications. For example, the above-described embodiments have been described in detail to clearly explain the present invention, and the present invention is not necessarily limited to those including all of the described configurations. Furthermore, it is possible to replace part of the configuration of one embodiment with the configuration of another embodiment, or to add the configuration of another embodiment to the configuration of one embodiment. Furthermore, it is possible to add, delete, or replace part of the configuration of each embodiment with other configurations.

[0090] Furthermore, the above-mentioned configurations, functions, processing units, processing means, etc. may be partly or entirely implemented in hardware, for example, by designing them as integrated circuits. The above-mentioned configurations, functions, etc. may also be implemented in software, with a processor interpreting and executing a program that implements each function. Information such as the programs, tables, and files that implement each function can be stored in a memory, a recording device such as a hard disk or SSD (Solid State Drive), or a recording medium such as an IC card, SD card, or DVD. In addition, the control lines and information lines shown are those that are considered necessary for the explanation, and do not necessarily show all the control lines and information lines in the product. In reality, it can be assumed that almost all components are interconnected. [Explanation of symbols]

[0091] 1 Clothing product suggestion device 2. User terminal equipment 3 External information device 4 Network 5. Storage 6. Washing machine 7. Clothes drying area 8 Cameras (RFID Readers) 11 Central control unit 12 Input Devices 13 Output Devices 14 Main memory 15 Auxiliary storage 16. Communications equipment 21 Input processing section 22 Clothing Proposal Department 23 Output Processing Section 31 Clothing information 32 Wearing history information 33 Proposed Model 34 Clothing Images (Clothing Images) 35 Coordination images (images of clothing combinations) 41 Proposal screen

Claims

1. an input processing unit that receives conditions regarding clothing and accessories worn by a user; a clothing accessory suggestion unit that determines, based on the received conditions, candidates for a clothing accessory combination that the user should wear from wearing history information that stores combinations of clothing accessories owned by the user that the user has actually worn in the past; and an output processing unit that displays the determined candidates on an output device; A clothing and accessories suggestion device comprising:

2. The clothing and accessories suggestion unit using a proposal model that takes the conditions as input and the candidates as output; The clothing and accessories suggestion device according to claim 1 .

3. The clothing and accessories suggestion unit machine learning the proposed model; The clothing and accessories suggestion device according to claim 2 .

4. The wearing history information is storing information for identifying an image of the clothing item; The output processing unit displaying the determined candidate image on an output device; The clothing and accessories suggestion device according to claim 1 .

5. The wearing history information is The evaluation of a combination of clothing accessories previously worn by the user is stored, The output processing unit determining an order in which the determined candidates are to be displayed on an output device based on the evaluation; The clothing and accessories suggestion device according to claim 1 .

6. The input processing unit of the clothing and accessories suggestion device includes: Accepting conditions regarding the clothing worn by the user; The clothing accessory suggestion unit of the clothing accessory suggestion device includes: determining, based on the received conditions, candidate combinations of clothing accessories to be worn by the user from wearing history information that stores combinations of clothing accessories owned by the user that the user has actually worn in the past; The output processing unit of the clothing and accessories suggestion device includes: displaying the determined candidates on an output device; The clothing suggestion method is characterized by the above.

Images