Computing device, method and apparatus for recommending at least one of a makeup palette and a hair dye color scheme - Patents.com

The computing device uses predictive models to generate objective color recommendations for makeup and hair dyes, addressing the challenge of subjective color matching, enabling harmonious and coordinated looks.

JP7577080B2Active Publication Date: 2024-11-01LOREAL SA
View PDF 2 Cites 0 Cited by

Patent Information

Application Number
JP2021576377
Authority / Receiving Office
JP · JP
Patent Type
Patents
Current Assignee / Owner
Filing Date
2019-08-21
Publication Date
2024-11-01
Estimated Expiration
2039-08-21

AI Technical Summary

Technical Problem

Consumers face difficulties in selecting harmonious color matches for multiple regions when choosing cosmetics or hair dyes, as existing methods rely on subjective recommendations from color experts.

Method used

A computing device and method that extracts color sets from digital images, using predictive models like CHAID decision trees or multivariate Gaussian map classifiers to generate objective color recommendations for makeup palettes and hair dye schemes, based on a sample training set of color harmonies.

Benefits of technology

Enables consumers to easily select optimal color schemes for multiple makeup and hair dyes, ensuring a color-coordinated natural look by providing objective and harmonious recommendations.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 0007577080000001
    Figure 0007577080000001
  • Figure 0007577080000002
    Figure 0007577080000002
  • Figure 0007577080000003
    Figure 0007577080000003
Patent Text Reader

Abstract

A method and apparatus for recommending at least one of a makeup palette or a hair dye color scheme is provided, the method including the steps of: extracting a color set for at least one region of a digital image associated with a user; generating color recommendation information regarding at least one of a makeup palette or a hair dye color scheme for at least two other regions of the digital image associated with the user based on one or more inputs indicating the extracted color set; and generating one or more instances of a virtual representation of the makeup palette recommendation or the hair dye color scheme recommendation based on the color recommendation information.
Need to check novelty before this filing date? Find Prior Art

Description

[Technical field]

[0001] The present disclosure relates to the field of beauty industry. More particularly, the present disclosure relates to a computing device, method, and apparatus for recommending at least one of a makeup palette and / or a hair dye color scheme. [Background technology]

[0002] Beauty has always been pursued by consumers. Researchers have established that an individual's appearance has a significant impact on one's self-confidence as well as the first impression of others. In particular, color matching is a focus when selecting cosmetics such as makeup products, hair dye products, and other cosmetics.

[0003] However, it is difficult for consumers to select the color they desire.

[0004] In the prior art, color experts such as makeup KOLs usually recommend color matching based on personal experience and color theory.However, such methods are subjective and only recommend harmonious color matching for each pair of regions, and therefore do not meet the demand for recommending harmonious color matching for more regions.

[0005] Therefore, there is a need to recommend makeup and / or hair dye colors for more areas to achieve a color-matched, natural look. Summary of the Invention [Means for solving the problem]

[0006] This Summary introduces a selection of concepts in a simplified form that are further described below in the Detailed Description. This Summary is not intended to identify key features or essential features of the claimed subject matter, nor is it intended to be used to limit the scope of the claimed subject matter.

[0007] Various aspects and features of the disclosure are described in further detail below.

[0008] According to a first aspect of the present disclosure, there is provided a computing device comprising: an optimum color unit including a computational circuit configured to extract a color set of at least one region of a digital image associated with a user and generate color recommendation information regarding at least one of a makeup palette or a hair dye color scheme for at least two other regions of the digital image associated with the user based on one or more inputs indicating the extracted color set; and a makeup and hair color unit including a computational circuit configured to generate one or more instances of a virtual representation of a makeup palette recommendation or a hair dye color scheme recommendation based on the color recommendation information.

[0009] In one embodiment of the first aspect of the present disclosure, the optimal color unit includes a computation circuit further configured to: store the extracted color set in a memory; arrange the extracted color set in a sample training set, the sample training set including information on colors of different regions in each of a plurality of sampled images, information on color schemes for different regions in each of the plurality of sampled images, and information on a total score for each color scheme; search for color schemes in the sample training set that include the extracted color set; rank total scores corresponding to the searched color schemes; and generate a recommended color scheme based on the ranked total scores.

[0010] In another embodiment of the first aspect of the present disclosure, the optimal color unit includes a computation circuit further configured to generate color recommendation information regarding at least one of a makeup palette or a hair dye color scheme for at least two other regions of the digital image associated with the user based on one or more inputs indicating a color set extracted by the predictive model. More specifically, the predictive model includes one of a Chi-squared Automatic Interaction Detector (CHAID) decision tree model, a multivariate Gaussian map classifier, and the like.

[0011] In yet another embodiment of the first aspect of the present disclosure, when the predictive model includes a Chi-squared Automatic Interaction Detector (CHAID) decision tree model, the optimal color unit includes a computation circuit further configured to: input the extracted color set into the CHAID decision tree model, the extracted color set being categorical, and the CHAID decision tree model being trained from a sample training set, the sample training set including information regarding colors of different regions in each of a plurality of sampled images, information regarding color schemes for different regions in each of the plurality of sampled images, and information regarding a total score for each color scheme; and generate color recommendation information regarding at least one of a makeup palette or a hair dye color scheme for at least two other regions of a digital image associated with a user.

[0012] In yet another embodiment of the first aspect of the present disclosure, when the predictive model includes a Chi-Square Automatic Interaction Detector (CHAID) decision tree model, the optimal color unit includes a computation circuit further configured to convert the extracted color set into corresponding color numerical values ​​in a color space; input the corresponding color numerical values ​​into the CHAID decision tree model, the CHAID decision tree model being trained from a sample training set, the sample training set including information on color numerical values ​​for different regions in each of a plurality of sampled images and information on a total score for each color scheme; and generate color recommendation information regarding at least one of a makeup palette or a hair dye color scheme for at least two other regions of a digital image associated with a user.

[0013] In yet another embodiment of the first aspect of the present disclosure, when the predictive model includes a multivariate Gaussian map classifier, the optimal color unit includes a computation circuit further configured to convert the extracted color set into corresponding color numerical values ​​in a color space; input the corresponding color numerical values ​​to the multivariate Gaussian map classifier, the multivariate Gaussian map classifier being trained from a sample training set, the sample training set including information on color numerical values ​​of different regions in each of a plurality of sampled images and information on a total score for each color scheme; and generate color recommendation information regarding at least one of a makeup palette or a hair dye color scheme for at least two other regions of a digital image associated with a user.

[0014] According to a second aspect of the present disclosure, there is provided a method for recommending at least one of a makeup palette or a hair dye color scheme, the method including the steps of extracting a color set for at least one region of a digital image associated with a user, generating color recommendation information regarding at least one of a makeup palette or a hair dye color scheme for at least two other regions of the digital image associated with the user based on one or more inputs indicating the extracted color set, and generating one or more instances of a virtual representation of the makeup palette recommendation or the hair dye color scheme recommendation based on the color recommendation information.

[0015] In one embodiment of the second aspect of the present disclosure, the step of generating color recommendation information regarding at least one of a makeup palette or a hair dye color scheme for at least two other regions of a digital image associated with a user based on one or more inputs indicating the extracted color set includes the steps of storing the extracted color set in a memory; arranging the extracted color set in a sample training set, the sample training set including information regarding colors of different regions in each of a plurality of sampled images, information regarding color schemes for different regions in each of the plurality of sampled images, and information regarding a total score for each color scheme; searching for color schemes in the sample training set including the extracted color set; ranking total scores corresponding to the searched color schemes; and generating a recommended color scheme based on the ranked total scores.

[0016] In another embodiment of the second aspect of the present disclosure, the step of generating color recommendation information regarding at least one of a makeup palette or a hair dye color scheme for at least two other regions of the digital image associated with the user based on one or more inputs indicating the extracted color set is realized by a predictive model. More specifically, the predictive model includes one of a Chi-squared Automatic Interaction Detector (CHAID) decision tree model, a multivariate Gaussian map classifier, and the like.

[0017] In another embodiment of the second aspect of the present disclosure, when the predictive model includes a Chi-Square Automatic Interaction Detector (CHAID) decision tree model, the step of generating color recommendation information regarding at least one of a makeup palette or a hair dye color scheme for at least two other regions of the digital image associated with the user based on one or more inputs indicating the extracted color set includes the steps of inputting the extracted color set into the CHAID decision tree model, the extracted color set being categorical, and the CHAID decision tree model being trained from a sample training set, the sample training set including information regarding colors of different regions in each of a plurality of sampled images, information regarding color schemes for different regions in each of the plurality of sampled images, and information regarding a total score for each color scheme; and generating color recommendation information regarding at least one of a makeup palette or a hair dye color scheme for the at least two other regions of the digital image associated with the user.

[0018] In yet another embodiment of the second aspect of the present disclosure, when the predictive model includes a Chi-squared Automatic Interaction Detector (CHAID) decision tree model, the step of generating color recommendation information regarding at least one of a makeup palette or a hair dyeing color scheme for at least two other regions of the digital image associated with the user based on one or more inputs indicating the extracted color set includes the steps of converting the extracted color set into corresponding color numerical values ​​in a color space; inputting the corresponding color numerical values ​​into the CHAID decision tree model, the CHAID decision tree model being trained from a sample training set, the sample training set including information regarding color numerical values ​​in a color space for different regions in each of a plurality of sampled images and information regarding a total score for each color scheme; and generating color recommendation information regarding at least one of a makeup palette or a hair dyeing color scheme for the at least two other regions of the digital image associated with the user.

[0019] In yet another embodiment of the second aspect of the present disclosure, when the predictive model includes a multivariate Gaussian map classifier, the step of generating color recommendation information regarding at least one of a makeup palette or a hair dyeing color scheme for at least two other regions of the digital image associated with the user based on one or more inputs indicating the extracted color set includes the steps of converting the extracted color set into corresponding color numerical values ​​in a color space; inputting the corresponding color numerical values ​​into the multivariate Gaussian map classifier, the multivariate Gaussian map classifier being trained from a sample training set, the sample training set including information regarding color numerical values ​​in a color space for different regions in each of a plurality of sampled images and information regarding a total score for each color scheme; and generating color recommendation information regarding at least one of a makeup palette or a hair dyeing color scheme for at least two other regions of the digital image associated with the user.

[0020] According to a third aspect of the present disclosure, there is provided an apparatus for recommending at least one of a makeup palette or a hair dye color scheme, the apparatus comprising: means for extracting a color set for at least one region of a digital image associated with a user; means for generating color recommendation information regarding at least one of a makeup palette or a hair dye color scheme for at least two other regions of the digital image associated with the user based on one or more inputs indicating the extracted color set; and means for generating one or more instances of a virtual representation of the makeup palette recommendation or the hair dye color scheme recommendation based on the color recommendation information.

[0021] According to a fourth aspect of the present disclosure, there is provided a computer-readable medium having stored thereon instructions which, when executed, cause a computing device to perform a method of recommending at least one of a makeup palette or a hair dye color scheme according to the second aspect of the present disclosure.

[0022] According to the present disclosure, the at least one region, the at least two other regions, and the respective distinct regions may be regions in which a body part of the user is located.

[0023] According to the present disclosure, the present invention allows consumers to easily select optimal color schemes for multiple makeup and / or hair dyes (more than two) simultaneously to achieve a color-harmonious and natural look. Furthermore, the present invention can further expand the ability to provide color advice to more consumers (including different ethnicities, groups, or individuals) by including new color options in the training set using the same methodology.

[0024] The above and other aspects, features, and advantages of various embodiments of the present disclosure will become more fully apparent from the following detailed description, taken in conjunction with the accompanying drawings, in which, by way of example, the same reference numbers or characters are used to indicate the same or equivalent elements. The drawings are presented to allow a better understanding of the embodiments of the present disclosure, but are not necessarily drawn to scale. [Brief description of the drawings]

[0025] [Figure 1] FIG. 2 is a block diagram of a computing device according to a first aspect of the present disclosure. [Diagram 2] 1 is a flowchart of a method for recommending at least one of a makeup palette or a hair dye color scheme according to a second embodiment of the present disclosure. [Diagram 3] FIG. 2 illustrates an exemplary sample training set according to the present disclosure. [Figure 4] 1 is a flowchart of a method for recommending at least one of a makeup palette or a hair dye color scheme according to an embodiment of the second aspect of the present disclosure. [Diagram 5] 13 is a flowchart of a method for recommending at least one of a makeup palette or a hair dye color scheme according to another embodiment of the second aspect of the present disclosure. [Figure 6] FIG. 1 illustrates an exemplary Chi-squared Automatic Interaction Detector (CHAID) decision tree model. [Figure 7] 13 is a flowchart of a method for recommending at least one of a makeup palette or a hair dye color scheme according to another embodiment of the second aspect of the present disclosure. [Figure 8] FIG. 1 illustrates another exemplary Chi-squared Automatic Interaction Detector (CHAID) decision tree model. [Figure 9] 13 is a flowchart of a method for recommending at least one of a makeup palette or a hair dye color scheme according to another embodiment of the second aspect of the present disclosure. [Figure 10]FIG. 11 is a block diagram of an apparatus for recommending at least one of a makeup palette or a hair dye color scheme according to a third embodiment of the present disclosure. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0026] The embodiments herein will be described in detail below with reference to the accompanying drawings illustrating the embodiments. However, these embodiments herein may be embodied in many different forms and should not be construed as being limited to the embodiments set forth herein. The elements of the drawings are not necessarily to scale relative to each other. Like reference characters refer to like elements throughout.

[0027] The terminology used herein is for the purpose of describing particular embodiments and is not intended to be limiting. As used herein, the singular forms "a," "an," and "the" are intended to include the plural unless the context clearly indicates otherwise. It is further to be understood that the terms "comprises," "comprising," "includes," and / or "including," as used herein, specify the presence of stated features, integers, steps, operations, elements, and / or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or groups thereof.

[0028] Unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood. It is further understood that the terms used herein should be interpreted to have a meaning consistent with the meaning in the context of this specification and related art, and are not to be interpreted in an idealized or overly formal sense unless expressly defined herein.

[0029] The present technology is described below with reference to block diagrams and / or flowcharts of methods, apparatus (systems), and / or computer programs according to the present embodiments. It should be understood that blocks of the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, may be implemented by computer program instructions. These computer program instructions may be supplied to a processor, controller, or control unit of a general purpose computer, special purpose computer, and / or other programmable data processing apparatus to produce machine means whereby the instructions, executed via the computer and / or other programmable data processing apparatus, produce means for performing the functions / acts specified in the block diagrams and / or flowchart blocks.

[0030] Accordingly, the technology may be embodied in hardware and / or software (including firmware, resident software, microcode, etc.). Furthermore, the technology may take the form of a computer usable or computer readable storage medium having computer usable or computer readable program code embodied in the medium for use by or in connection with an instruction execution system. In the context of this specification, a computer usable or computer readable medium may be any means that may contain or store, communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device.

[0031] Hereinafter, embodiments of the present invention will be described with reference to the drawings.

[0032] FIG. 1 is a block diagram of a computing device 100 according to a first aspect of the present disclosure.

[0033] 1, the computing device 100 includes an optimum color unit 101 and a makeup and hair color unit 102 coupled together. The optimum color unit 101 includes computation circuitry configured to perform the following operations:

[0034] In the present disclosure, in order to obtain optimal color schemes for a number of makeups and / or hair dyes, a digital image of a user is obtained by a device such as a camera, a handset having a camera, etc. In one example, such digital image can be obtained by uploading an existing image to an App. A color set of at least one region of the relevant digital image associated with the user is extracted by the computation circuit included in the optimal color unit 101. In one example, the region can be a region where a part of a human body is located. The extracted color set can be one or more inputs. Then, color recommendation information regarding at least one of a makeup palette or a hair dye color scheme for at least two other regions of the digital image associated with the user based on one or more inputs indicating the extracted color set is generated by ranking a predictive model or a sample training set. Details on how to generate color recommendation information regarding at least one of a makeup palette or a hair dye color scheme for at least two other regions of the digital image associated with the user based on one or more inputs indicating the extracted color set are described below. Finally, the recommended colors of makeup and / or hair dye are generated. The form of the generated recommended colors of makeup and / or hair dye can be in various forms. In one example, the makeup and hair color unit 102 includes computational circuitry configured to generate one or more instances of a virtual representation of a makeup palette recommendation or a hair color scheme recommendation based on the color recommendation information.

[0035] Computing device 100 may be, for example, a service provider's server, a client-related device (e.g., a client device), a system-on-a-chip, and / or any other suitable computing device or computing system. In various implementations, computing device 100 may take on a variety of different configurations. For example, computing device 100 may be implemented as a computer-like device, including a personal computer, a desktop computer, a multi-screen computer, a laptop computer, a netbook, and the like. Computing device 100 may also be implemented as a mobile device-like device, including mobile devices such as a mobile phone, a portable music player, a portable gaming device, a tablet computer, a multi-screen computer, and the like. Computing device 100 may also be implemented as a television-like device having or connected to an overall large screen in a typical viewing environment. These devices include televisions, set-top boxes, game consoles, and the like.

[0036] In one embodiment, the computational circuitry includes one or more computing devices, such as, among others, a processor (e.g., a microprocessor), a central processing unit (CPU), a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA), or the like, or any combination thereof, and may include discrete digital or analog circuit elements or electronics, or combinations thereof. In one embodiment, the computational circuitry includes one or more ASICs having multiple predefined logic components. In one embodiment, the computational circuitry includes one or more FPGAs, each having multiple programmable logic components.

[0037] In one embodiment, the computational circuitry includes one or more electrical circuits, printed circuits, flexible circuits, electrical conductors, electrodes, cavity resonators, conducting traces, ceramic patterned electrodes, electromechanical components, transducers, and the like.

[0038] In one embodiment, the computational circuitry includes one or more components operably coupled (e.g., communicatively, electromagnetically, magnetically, ultrasonically, optically, inductively, electrically, capacitively coupled, wirelessly coupled, etc.) to one another. In one embodiment, the circuitry includes one or more remotely located components. In one embodiment, the remotely located components are operably coupled, for example, via wireless communication. In one embodiment, the remotely located components are operably coupled, for example, via one or more communication modules, receivers, transmitters, transceivers, etc.

[0039] In one embodiment, the computation circuit includes a memory that stores, for example, instructions or information. Non-limiting examples of memory include volatile memory (e.g., random access memory (RAM), dynamic random access memory (DRAM), etc.), non-volatile memory (e.g., read only memory (ROM), electrically erasable programmable read only memory (EEPROM), compact disc read only memory (DC-ROM), etc.), persistent memory, etc. Further non-limiting examples of memory include erasable programmable read only memory (EPROM), flash memory, etc. In one embodiment, the memory is coupled to one or more computing devices, for example, by one or more instruction, information, or power buses. In one embodiment, the computation circuit includes one or more databases stored in the memory. In one embodiment, the computation circuit includes one or more look-up tables stored in the memory.

[0040] In one embodiment, the computational circuitry includes one or more computer readable medium drives, interface sockets, Universal Serial Bus (USB) ports, memory card slots, etc., and one or more input / output components, such as, for example, a graphical user interface, a display, a keyboard, a keypad, a trackball, a joystick, a touch screen, a mouse, switches, dials, and any other peripheral devices. In one embodiment, the computational circuitry includes one or more user input / output components operably coupled to at least one computing device configured to, for example, control (electrical, electromechanical, software-implemented, firmware-implemented, or other control, or a combination thereof) at least one parameter associated with determining a thermal property of one or more tissues in response to detecting a shift in the turn-on voltage.

[0041] In one embodiment, the computational circuitry includes electrical circuitry forming a general-purpose computing device configured by a computer program (e.g., a general-purpose computer configured by a computer program to at least partially execute the processes and / or devices described herein, or a microprocessor configured by a computer program to at least partially execute the processes and / or devices described herein), electrical circuitry forming a memory device (e.g., a form of memory (e.g., random access, flash, read-only, etc.)), electrical circuitry forming a communications device (e.g., a modem, a communications switch, an optical electrical appliance, etc.), and / or any non-electrical analog or other analog thereto, such as optical.

[0042] The computing device 100 according to the first aspect of the present disclosure enables consumers to easily select optimal color schemes for multiple makeup and / or hair dyes (more than two) simultaneously to achieve a color-harmonious and natural look.

[0043] FIG. 2 is a flow chart of a method 200 for recommending at least one of a makeup palette and a hair dye color scheme according to a second aspect of the present disclosure.

[0044] This method can overcome the above and other potential deficiencies in the conventional approach. Method 200 starts at block 201. In block 201, a color set of at least one region of a digital image associated with a user is extracted. In one example, the digital image associated with the user is obtained by a device such as a camera, a handset with a camera, etc. In one example, such digital image can be obtained by uploading an existing image to an App. Furthermore, the digital image can be an image having a raw color or a makeup color. In one example, the region can be an area where a part of a human body is located. In one embodiment of the present disclosure, the at least one region can be an area where skin, hair, lips, eyebrows, makeup skin / hair / lips / eyebrows, clothing, fingernails, shoes, accessories, etc. are located. Of course, the present disclosure is not limited to such regions and can be any region of a human body without departing from the spirit and scope of the present disclosure. If a consumer wants to fix the color of two or more regions (e.g., leave the skin color as well as the hair color intact), all these fixed regions are measured. In block 202, color recommendation information is generated regarding at least one of a makeup palette or a hair color scheme for at least two other regions of the digital image associated with the user based on one or more inputs indicating the extracted color set. Finally, in block 203, the recommended colors are generated. More specifically, block 203 includes generating one or more instances of a virtual representation of a makeup palette recommendation or a hair color scheme recommendation based on the color recommendation information. In the following, it will be introduced in detail how the makeup and hair color unit 102 generates one or more instances of a virtual representation of a makeup palette recommendation or a hair color scheme recommendation based on the color recommendation information.

[0045] Thus, the present invention allows consumers to easily select optimal color schemes for multiple makeup and / or hair dyes (more than two) simultaneously to produce a color-matched, natural look.

[0046] FIG. 3 illustrates an exemplary sample training set 300 according to this disclosure.

[0047] An exemplary sample training set 300 is created based on 200 users. The four most common and important parts of women are selected, namely, hair tone, skin tone, lip color, and eyebrow color. For each part, we use the most common color for consumers, and then combine the colors one by one to get a total of 4*4*3*2=96 different looks as follows: 4 hair colors (top 4 hair colors according to consumers from hair topology, 90% of all samples) - Black - Dark brown - brown - Light brown 4 Skin Tones - (From the skin topology, select 4 out of 7 skin tones that are more representative of makeup consumers aged 18-40 years old for different skin lightness and hue) - "Fair" - bright yellow - "Pink" - bright red - "Beige" - a darker yellow - "Dark Red" - a darker red color 3 Lip Shades - MBL Lipsticks - (generally lipstick colors from the three basic shade zones, i.e. red, orange, and pink) - Red-MBL R09pm - Orange-MBL C06 - Pink-MBL P16 Two eyebrow colors - (brown, the most common on the market) - Grey - brown

[0048] A three-scale scoring system was developed for consumers' perception of "harmony". The system randomly displays each of the 96 photos on the screen, and the consumer answers "harmonious", "neutral", or "not harmonious" to the question "Is the appearance of the color match harmonious?", and the system records a score of 1, 0, or -1, respectively. Data on each appearance is recorded for each consumer. Then, a total score for each appearance is calculated, which is the final "harmony score". Finally, a sample training set 300 is created as shown in FIG. 3. Such a sample training set 300 includes information on the color of the different regions of each of the multiple sampled images, information on the color scheme for the different regions of each of the multiple sampled images, and information on the total score for each color scheme as described in the sample training set 300. For example, the first row shows light brown-brown-beige-pink as the color scheme, light brown, brown, beige, pink as the colors, and -34 as the total harmony score. Such a sample training set is very important in implementing the present invention. The information in such a sample training set is not limited to this. Such sample training set 300 is merely exemplary, and the present disclosure is not limited to such sample training set. As mentioned above, such sample training sets can also be created based on skin, hair, lips, eyebrows, makeup skin / hair / lips / eyebrows, clothing, fingernails, shoes, accessories, etc., and associated total scores for each color scheme.

[0049] FIG. 4 shows a flow chart of a method 400 for recommending at least one of a makeup palette and a hair dye color scheme according to an embodiment of the second aspect of the present disclosure.

[0050] This method can overcome the above and other potential deficiencies in the conventional approach. The method includes blocks 401, 402, 403, 404, 405, 406, and 407. Block 401 in Fig. 4 is the same as block 201 in Fig. 2, and block 407 in Fig. 4 is the same as block 203 in Fig. 2. To avoid repetition, blocks 401 and 407 will not be described, and blocks 402, 403, 404, 405, and 406 will be described in detail.

[0051] In block 402, the extracted color set is stored in a memory. In block 403, the extracted color set is placed in a sample training set. The sample training set can be pre-created based on a number of sample users as shown in FIG. 3 and includes information on colors of different regions in each of a plurality of sampled images, information on color schemes for different regions in each of the plurality of sampled images, and information on a total score for each color scheme.

[0052] In block 404, color schemes that include the extracted color set are searched for in the sample training set. In block 405, corresponding total scores corresponding to the searched color schemes are ranked. In block 406, a recommended color scheme is generated based on the ranked total scores.

[0053] In one example, the device acquires a digital image including lips, extracts lip color as RED, and searches for color schemes including RED in the sample training set, and then ranks the total scores corresponding to the searched color schemes including RED. Finally, a recommended color scheme based on the ranked total scores is generated to be presented to a user. In another example, the device can acquire a digital image including skin, hair, lips, eyebrows, makeup skin / hair / lips / eyebrows, clothing, fingernails, shoes, accessories, etc. Then, makeup and / or hair dye recommended colors for at least two other regions of the source image are generated based on the extracted color set. Finally, a recommended color scheme according to the present disclosure is presented to a user. In the present disclosure, one or more examples of a virtual representation of a makeup palette recommendation or a hair dye color scheme recommendation based on color recommendation information can be presented to a user, as shown in block 407.

[0054] The method 400 for recommending a makeup palette and / or hair dye color scheme according to an embodiment of the second aspect of the present disclosure can provide color harmony for an overall look.

[0055] In an embodiment of the second aspect, the block 203 of generating color recommendation information regarding at least one of a makeup palette or a hair dye color scheme for at least two other regions of the digital image associated with the user based on one or more inputs indicating the extracted color set is realized by a predictive model, which includes one of a Chi-squared Automatic Interaction Detector (CHAID) decision tree model, a multivariate Gaussian map classifier, etc., which are described in more detail below.

[0056] FIG. 5 is a flow chart of a method 500 for recommending at least one of a makeup palette and a hair dye color scheme according to another embodiment of the second aspect of the present disclosure.

[0057] This method can overcome the above and other potential deficiencies in the conventional approach. The method includes blocks 501, 502, 503, and 504. Block 501 in Fig. 5 is the same as block 201 in Fig. 2, and block 504 in Fig. 4 is the same as block 203 in Fig. 2. To avoid repetition, blocks 501 and 504 will not be described, and blocks 502 and 503 will be described in detail.

[0058] In one embodiment, the predictive model may include a Chi-Square Automatic Interaction Detector (CHAID) decision tree model that creates a non-binary tree (i.e., a tree in which three or more branches can be attached to a single root or node) based on a relatively simple algorithm that is particularly suitable for analyzing larger data sets. In this case, in block 502, the extracted color set is input into the CHAID decision tree model. The extracted color set is a categorical type, such as red, black, orange, etc., and the CHAID decision tree model is trained from a sample training set, the sample training set including information on the colors of different regions in each of a plurality of sampled images, information on color schemes for different regions in each of the plurality of sampled images, and information on a total score for each color scheme. Such a sample training set is shown in FIG. 3. To make such a model more accurate, after creating the CHAID decision tree model, it is necessary to validate the CHAID decision tree model. Then, in block 503, color recommendation information is generated regarding at least one of a makeup palette or a hair dye color scheme for at least two other regions of the digital image associated with the user. Finally, as shown in block 504, the user may be presented with one or more examples of virtual representations of makeup palette recommendations or hair dye color scheme recommendations based on the color recommendation information.

[0059] FIG. 6 shows an exemplary Chi-squared Automatic Interaction Detector (CHAID) decision tree model 600. FIG. 6 is an exemplary CHAID decision tree model in which lips are taken as the starting node. However, with respect to the starting point of the decision tree, the color scheme introduced by each case may lead to different starting points for each of the decision trees, i.e., in some cases, lips are the most likely, and in other cases, hair, skin, or eyebrows are the most likely. In this exemplary case, if we want to determine skin color first, we can look for lip color and other matches, and then create a partial decision tree for each skin color, so we must first determine skin color and then enter the partial decision tree. That is, if we want to start the decision tree by identifying skin tone, we may create one partial decision tree for each skin tone by the same algorithm.

[0060] The exemplary Chi-squared Automatic Interaction Detector (CHAID) decision tree model 600 presents a greater range of harmonious colors to consumers.

[0061] The exemplary Chi-squared Automatic Interaction Detector (CHAID) decision tree model 600 is merely exemplary and can be created according to different needs. For example, if the starting node is at least one of skin, hair, lips, eyebrows, makeup skin / hair / lips / eyebrows, clothing, fingernails, shoes, accessories, etc., makeup and / or hair dye color recommendations for at least two other areas are generated. Finally, the recommended color scheme is presented to the consumer.

[0062] FIG. 7 is a flowchart of a method 700 for recommending at least one of a makeup palette and a hair dye color scheme according to another embodiment of the second aspect of the present disclosure.

[0063] This method can overcome the above and other potential deficiencies in the conventional approaches. Method 700 includes blocks 701, 702, 703, 704, and 705. Block 701 in Fig. 7 is the same as block 201 in Fig. 2, and block 705 in Fig. 7 is the same as block 203 in Fig. 2. To avoid repetition, blocks 701 and 705 will not be described, and blocks 702, 703, and 704 will be described in detail.

[0064] If the prediction model can include a Chi-squared Automatic Interaction Detector (CHAID) decision tree model, the extracted color set can be expressed as a numerical value of the corresponding color in a color space. For this purpose, in block 702, the extracted color set is converted into a numerical value of the corresponding color in a color space. Such numerical values ​​of the colors can be numerical values ​​in several color spaces:

[0065] A color space as described above may be defined as a multi-dimensional space, where points in the multi-dimensional space represent lightness and the dimensions of the space represent variables in the color model. For example, in a red, green, blue (RGB) color space, an additive color model defines three variables, each representing a different amount of red, green, and blue light. In a digital model, the values ​​of these quantities may be defined with reference to a quantized set of values. For example, a color defined using an 8-bit RGB model may have three values ​​stored in memory, where each variable may be assigned a value between 0 and 255. Other colour spaces include the cyan, magenta, yellow, and black (CMYK) colour space, which uses four variables to represent different amounts of each colourant (for example in a printing system) in a subtractive colour model; the Commission Internationale de Illumination (CIE) 1931 XYZ colour space, which models colour using three variables ('X', 'Y' and 'Z' or tristimulus values); the CIE 1976 (L*, a*, b*-CIELAB or 'LAB') colour space, in which three variables represent lightness ('L') and opposing colour dimensions ('a' and 'b'); and the Yu'v' colour space, in which three variables represent luminance ('Y') and two chrominance dimensions (u' and v').

[0066] In block 703, the corresponding numerical values ​​of the inputs are input to the CHAID decision tree model. The CHAID decision tree model is trained from a sample training set, which includes information about numerical values ​​of colors in a color space for different regions in each of a plurality of sampled images and information about a total score for each color scheme. Such a sample training set can be created in advance. To make such a model more accurate, after creating the CHAID decision tree model, it is necessary to validate the CHAID decision tree model.

[0067] In block 704, color recommendation information is generated regarding at least one of a makeup palette or a hair color scheme for at least two other regions of the digital image associated with the user. Finally, the color recommendation information can be presented to the user. In one example, one form of color recommendation information can be one or more instances of a virtual representation of a makeup palette recommendation or a hair color scheme recommendation, as shown in block 805.

[0068] Method 700 can also create a decision tree model according to the color value of each region, and then determine whether the harmony is based on color value. At the same time, it can predict whether a certain color combination is good within a certain range.

[0069] FIG. 8 illustrates another exemplary Chi-Square Automatic Interaction Detector (CHAID) decision tree model 800.

[0070] In this Chi-squared Automatic Interaction Detector (CHAID) decision tree model, the input is the numerical value of color (I, a, b). By using the numerical value of color (I, a, b), a recommendation decision tree model for prediction can be created, which not only recommends the exact tested color combination in the database, but also allows predicting some other colors close to the desired range. Of course, the color space can be extended to the above-mentioned color space without departing from the scope of protection of the present invention.

[0071] FIG. 9 is a flowchart of a method 900 for recommending at least one of a makeup palette and a hair dye color scheme according to another embodiment of the second aspect of the present disclosure.

[0072] This method can overcome the above and other potential deficiencies in the conventional approach. The method 900 includes blocks 901, 902, 903, 904, and 905. Block 901 in Fig. 9 is the same as block 201 in Fig. 2, and block 905 in Fig. 9 is the same as block 203 in Fig. 2. To avoid repetition, blocks 901 and 905 will not be described, and blocks 902, 903, and 904 will be described in detail.

[0073] When the predictive model includes a multivariate Gaussian map classifier, the extracted color set is converted to corresponding color values ​​in a color space in block 902. Such color values ​​can be in several color spaces mentioned above. In block 903, the corresponding color values ​​are input to a multivariate Gaussian map classifier. The multivariate Gaussian map classifier is trained from a sample training set, which includes information on color values ​​for different regions in each of a plurality of sampled images and information on the total score for each color combination.

[0074] In block 904, color recommendation information is generated for at least one of a makeup palette or a hair color scheme for at least two other regions of the digital image associated with the user based on one or more inputs indicating the extracted color set. In this method 900, by inputting a new data point into the space, it is determined whether the data point is "close enough" to any of the data clouds to predict whether a color scheme is harmonious or not. Finally, color recommendation information can be presented to the user. In one example, one form of color recommendation information can be one or more instances of a virtual representation of a makeup palette recommendation or a hair color scheme recommendation as shown in block 905.

[0075] FIG. 10 shows a block diagram of an apparatus 1000 for recommending at least one of a makeup palette or a hair dye color scheme according to a third embodiment of the present disclosure.

[0076] The apparatus 100 for recommending at least one of a makeup palette or a hair dyeing color scheme comprises means 1001 for extracting a color set of at least one region of a digital image associated with a user, means 1002 for generating color recommendation information regarding at least one of a makeup palette or a hair dyeing color scheme for at least two other regions of the digital image associated with the user based on one or more inputs indicating the extracted color set, and means 1003 for generating one or more instances of a virtual representation of the makeup palette recommendation or hair dyeing color scheme recommendation based on the color recommendation information.

[0077] The means 1002 for generating color recommendation information regarding at least one of a makeup palette or a hair dye color scheme for at least two other regions of the digital image associated with the user based on one or more inputs indicating the extracted color set may be realized by a predictive model or a computational circuit.

[0078] The computing circuitry may include means for storing the extracted color set in a memory; means for arranging the extracted color set into a sample training set, the sample training set including information about colors of different regions in each of a plurality of sampled images, information about color schemes for different regions in each of the plurality of sampled images, and information about a total score for each color scheme; means for searching the sample training set for color schemes that include the extracted color set; means for ranking total scores corresponding to the searched color schemes; and means for generating a recommended color scheme based on the ranked total scores.

[0079] The predictive model may include, but is not limited to, a Chi-Square Automatic Interaction Detector (CHAID) decision tree model or a multivariate Gaussian map classifier as described above.

[0080] The above description is merely illustrative and not limiting. The present disclosure can realize a technical solution, and when at least one color set of skin, hair, lips, eyebrows, makeup skin / hair / lips / eyebrows, clothing, fingernails, shoes, accessories, etc. is input, at least two recommended colors of makeup skin / hair / lips / eyebrows can be output.

[0081] According to the above disclosure, the present invention can easily simultaneously recommend optimal color schemes for multiple makeup or hair dyes (more than two) based on a scientific and objective evaluation of the consumer's color parameters (natural skin tone, hair color, eye color, etc.), resulting in a color-harmonious and natural look.

[0082] An embodiment of the present disclosure may be an article of manufacture having stored on a non-transitory machine-readable medium (such as a microelectric memory) instructions (e.g., computer code) that program one or more processing components (generically referred to herein as "processors") to perform the operations described above. In other embodiments, some of these operations may be performed by specific hardware components that include hard-wired logic (e.g., dedicated digital filter blocks and state machines). Alternatively, the operations may be performed by any combination of programmed data processing components and hard-wired circuit components.

[0083] Although the embodiments have been shown and described herein, it will be understood by those skilled in the art that various changes and modifications may be made and that elements of the embodiments may be substituted with equivalents without departing from the true scope of the present technology. Furthermore, many modifications may be made to adapt to a particular situation and the teachings of the present specification without departing from the central scope of the teachings of the present specification. Therefore, it is not intended that the present embodiments be limited to the particular embodiment disclosed as the best mode contemplated for carrying out the present technology, but rather that the present embodiments include all embodiments within the scope of the appended claims. [Explanation of symbols]

[0084] 100 computing devices 101 Optimal Color Unit 102 Makeup and Hair Color Unit 300 illustrative sample training set 600, 800 Chi-squared Automatic Interaction Detector (CHAID) decision tree model 1000 devices

Claims

1. 1. A computing device comprising: an optimization color unit including a computation circuit configured to extract a color set for at least one region of a digital image associated with a user, and generate color recommendation information regarding at least one of a makeup palette or a hair dye color scheme for at least two other regions of the digital image associated with the user based on one or more inputs indicative of the extracted color set; a makeup and hair color unit including a computational circuit configured to generate one or more instances of a virtual representation of a makeup palette recommendation or a hair dye color combination recommendation based on the color recommendation information; the color recommendation information being a recommendation for a color combination of the at least one region and the at least two other regions; The optimal color unit is storing the extracted color set in a memory; arranging the extracted color set into a sample training set, the sample training set including information about colors of different regions in each of a plurality of sampled images, information about color schemes for different regions in each of the plurality of sampled images, and information about a total score for each color scheme; Searching for color schemes including the extracted color set in the sample training set; ranking the total scores corresponding to the searched color schemes; and generating a recommended color scheme based on the ranked total score. Computing device.

2. The computing device of claim 1, wherein the optimal color unit further comprises a computation circuit configured to generate color recommendation information regarding at least one of a makeup palette or a hair dye color scheme for at least two other regions of the digital image associated with the user based on one or more inputs indicative of the color set extracted by the predictive model.

3. 3. The computing device of claim 2, wherein the predictive model comprises one of a Chi-Square Automatic Interaction Detector (CHAID) decision tree model and a multivariate Gaussian map classifier.

4. When the prediction model includes a Chi-squared Automatic Interaction Detector (CHAID) decision tree model, the optimal color unit is inputting the extracted color set into the CHAID decision tree model, the extracted color set being a category type, the CHAID decision tree model being trained from a sample training set, the sample training set including information regarding colors of different regions in each of a plurality of sampled images, information regarding color schemes for different regions in each of the plurality of sampled images, and information regarding a total score for each color scheme; and generating color recommendation information regarding at least one of a makeup palette or a hair dye color scheme for at least two other regions of the digital image associated with the user.

5. The predictive model includes a Chi-Square Automatic Interaction Detector (CHAID) decision tree model, and the optimal color unit is converting the extracted color set into numerical values ​​of corresponding colors in a color space; inputting the corresponding color values ​​into the CHAID decision tree model, the CHAID decision tree model being trained from a sample training set, the sample training set including information about color values ​​in a color space for different regions of each of a plurality of sampled images and information about a total score for each color scheme; and generating color recommendation information regarding at least one of a makeup palette or a hair dye color scheme for at least two other regions of the digital image associated with the user.

6. When the predictive model includes a multivariate Gaussian map classifier, the optimal color unit is converting the extracted color set into numerical values ​​of corresponding colors in a color space; inputting the corresponding color numerical values ​​into the multivariate Gaussian map classifier, the multivariate Gaussian map classifier being trained from a sample training set, the sample training set including information about color numerical values ​​in a color space for different regions in each of a plurality of sampled images and information about a total score for each color scheme; and generating color recommendation information regarding at least one of a makeup palette or a hair dye color scheme for at least two other regions of the digital image associated with the user.

7. 7. A computing device as claimed in claim 1, wherein the at least one region, the at least two other regions, and the respective distinct regions may be regions in which a part of the user's body is located.

8. 1. A method for recommending at least one of a makeup palette or a hair dye color scheme, comprising: Extracting a color set for at least one region of a digital image associated with a user; generating color recommendation information regarding at least one of a makeup palette or a hair dye color scheme for at least two other regions of the digital image associated with the user based on one or more inputs indicating the extracted color set; generating one or more instances of a virtual representation of a makeup palette recommendation or a hair dye color combination recommendation based on the color recommendation information; the color recommendation information being a recommendation for a color combination of the at least one region and the at least two other regions; generating color recommendation information regarding at least one of a makeup palette or a hair dye color scheme for at least two other regions of the digital image associated with the user based on one or more inputs indicating the extracted color set, storing the extracted color set in a memory; arranging the extracted color set into a sample training set, the sample training set including information about colors of different regions in each of a plurality of sampled images, information about color schemes for different regions in each of the plurality of sampled images, and information about a total score for each color scheme; searching for color schemes in the sample training set that include the extracted color set; ranking the total scores corresponding to the searched color schemes; and generating a recommended color scheme based on the ranked total scores. method.

9. 9. The method of claim 8, wherein the step of generating color recommendation information regarding at least one of a makeup palette or a hair dye color scheme for at least two other regions of the digital image associated with the user based on one or more inputs indicative of the extracted color set is accomplished by a predictive model.

10. 10. The method of claim 9, wherein the predictive model comprises one of a Chi-Square Automatic Interaction Detector (CHAID) decision tree model and a multivariate Gaussian map classifier.

11. When the predictive model includes a Chi-squared Automatic Interaction Detector (CHAID) decision tree model, the step of generating color recommendation information regarding at least one of a makeup palette or a hair dye color scheme for at least two other regions of the digital image associated with the user based on one or more inputs indicating the extracted color set includes: inputting the extracted color set into the CHAID decision tree model, the extracted color set being a category type, and the CHAID decision tree model being trained from a sample training set, the sample training set including information regarding colors of different regions in each of a plurality of sampled images, information regarding color schemes for different regions in each of the plurality of sampled images, and information regarding a total score for each color scheme; and generating color recommendation information regarding at least one of a makeup palette or a hair dye color scheme for at least two other regions of the digital image associated with the user.

12. When the predictive model includes a Chi-squared Automatic Interaction Detector (CHAID) decision tree model, the step of generating color recommendation information regarding at least one of a makeup palette or a hair dye color scheme for at least two other regions of the digital image associated with the user based on one or more inputs indicating the extracted color set includes: converting the extracted color set into numerical values ​​of corresponding colors in a color space; inputting the corresponding color numerical values ​​into the CHAID decision tree model, the CHAID decision tree model being trained from a sample training set, the sample training set including information regarding color numerical values ​​in a color space for different regions in each of a plurality of sampled images and information regarding a total score for each color scheme; and generating color recommendation information regarding at least one of a makeup palette or a hair dye color scheme for at least two other regions of the digital image associated with the user.

13. When the predictive model includes a multivariate Gaussian map classifier, the step of generating color recommendation information regarding at least one of a makeup palette or a hair dye color scheme for at least two other regions of the digital image associated with the user based on one or more inputs indicating the extracted color set includes: converting the extracted color set into numerical values ​​of corresponding colors in a color space; inputting the corresponding color numerical values ​​into the multivariate Gaussian map classifier, the multivariate Gaussian map classifier being trained from a sample training set, the sample training set including information regarding color numerical values ​​in a color space for different regions in each of a plurality of sampled images and information regarding a total score for each color scheme; and generating color recommendation information regarding at least one of a makeup palette or a hair dye color scheme for at least two other regions of the digital image associated with the user.

14. The method according to any one of claims 8, 11-13, wherein the at least one region, the at least two other regions and the respective distinct regions may be regions in which a part of the user's body is located.

15. A device for recommending at least one of a makeup palette and a hair dye color scheme, comprising: means for extracting a color set for at least one region of a digital image associated with a user; means for generating color recommendation information regarding at least one of a makeup palette or a hair dye color scheme for at least two other regions of the digital image associated with the user based on one or more inputs indicating the extracted color set; and means for generating one or more instances of a virtual representation of a makeup palette recommendation or a hair dye color combination recommendation based on the color recommendation information; the color recommendation information being a recommendation for a color combination of the at least one region and the at least two other regions; The means for generating color recommendation information regarding at least one of a makeup palette or a hair dye color scheme for at least two other regions of the digital image associated with the user based on one or more inputs indicating the extracted color set includes: means for storing the extracted color set; means for arranging the extracted color set into a sample training set, the sample training set including information regarding colors of different regions in each of a plurality of sampled images, information regarding color schemes for different regions in each of the plurality of sampled images, and information regarding a total score for each color scheme; A means for searching color schemes including the extracted color set in the sample training set; means for ranking the total scores corresponding to the searched color schemes; and means for generating a recommended color scheme based on the ranked total score. Device.

16. 16. The apparatus of claim 15, wherein the means for generating color recommendation information regarding at least one of a makeup palette or a hair dye color scheme for at least two other regions of the digital image associated with the user based on one or more inputs indicative of the extracted color set is realized by a predictive model.

17. 17. The apparatus of claim 16, wherein the predictive model comprises one of a Chi-Square Automatic Interaction Detector (CHAID) decision tree model and a multivariate Gaussian map classifier.

18. When the predictive model includes a Chi-squared Automatic Interaction Detector (CHAID) decision tree model, the means for generating color recommendation information regarding at least one of a makeup palette or a hair dye color scheme for at least two other regions of the digital image associated with the user based on one or more inputs indicating the extracted color set includes: a means for inputting the extracted color set into the CHAID decision tree model, the extracted color set being a category type, and the CHAID decision tree model being trained from a sample training set, the sample training set including information regarding colors of different regions in each of a plurality of sampled images, information regarding color schemes for different regions in each of the plurality of sampled images, and information regarding a total score for each color scheme; and generating color recommendation information regarding at least one of a makeup palette or a hair dye color scheme for at least two other regions of the digital image associated with the user.

19. When the predictive model includes a Chi-squared Automatic Interaction Detector (CHAID) decision tree model, the means for generating color recommendation information regarding at least one of a makeup palette or a hair dye color scheme for at least two other regions of the digital image associated with the user based on one or more inputs indicating the extracted color set includes: means for converting the extracted color set into numerical values ​​of corresponding colors in a color space; means for inputting the corresponding color values ​​into the CHAID decision tree model, the CHAID decision tree model being trained from a sample training set, the sample training set including information regarding color values ​​in a color space for different regions of each of a plurality of sampled images and information regarding a total score for each color scheme; and generating color recommendation information regarding at least one of a makeup palette or a hair dye color scheme for at least two other regions of the digital image associated with the user.

20. When the predictive model includes a multivariate Gaussian map classifier, the means for generating color recommendation information regarding at least one of a makeup palette or a hair dye color scheme for at least two other regions of the digital image associated with the user based on one or more inputs indicating the extracted color set includes: means for converting the extracted color set into numerical values ​​of corresponding colors in a color space; means for inputting the corresponding numerical color values ​​into the multivariate Gaussian map classifier, the multivariate Gaussian map classifier being trained from a sample training set, the sample training set including information regarding numerical color values ​​in a color space for different regions of each of a plurality of sampled images and information regarding a total score for each color scheme; and generating color recommendation information regarding at least one of a makeup palette or a hair dye color scheme for at least two other regions of the digital image associated with the user.

21. 21. The apparatus of claim 15, 18-20, wherein the at least one region, the at least two other regions, and the respective distinct regions may be regions in which a part of the user's body is located.

22. A computer-readable storage medium having stored thereon instructions for causing a computing device to perform the method of any one of claims 8 to 14.

Citation Information

Patent Citations

  • Virtual Makeover

    JP2003504131A

  • Makeup advice device, the makeup advice method and program

    JP2010211308A