A digital imaging and artificial intelligence-based system and method for analyzing pixel data of a user's skin image to generate one or more user-specific skin spot classifications.

Abstract

This paper describes a digital imaging and artificial intelligence-based system and method for analyzing pixel data from a user's skin image and generating one or more user-specific skin spot classifications. A user's digital image is received by an imaging application (app) and includes pixel data from at least a portion of the user's skin area. A skin-based learning model, trained using pixel data from multiple training images depicting each individual's skin, analyzes this image and determines at least one spot classification on the user's skin. Based on this at least one spot classification, the imaging app generates user-specific skin recommendations designed to address at least one identifiable spot feature within the pixel data containing a portion of the user's skin area.

Description

Technical Field

[0001] The present disclosure generally relates to digital imaging and artificial intelligence-based systems and methods, and more specifically, to digital imaging and artificial intelligence-based systems and methods for analyzing pixel data of an image of a user's skin to generate one or more user-specific skin blemish classifications.

Background Art

[0002] Generally, human skin can experience changes and / or discoloration in certain areas such as the face. The changes and / or discoloration can occur in the form of one or more blemishes on the skin surface in various manners. Such blemishes can be of various types, shapes, and sizes, shades, and / or colors. For example, one type of blemish can be a red blemish that can be a hemoglobin-related blemish. Hemoglobin-related blemishes can occur when blood accumulates under the skin and leaves residues of hemoglobin that settle in the tissue there. Hemoglobin contains iron, which causes a red or rust-colored skin tone in the affected area. As another example, another type of blemish can be a brown blemish that can be a melanin-related blemish. Generally, melanin is a substance that can increase pigmentation of the skin. Melanin in the skin can depend on various factors including genetic nature and sun exposure. Generally, skin blemishes, as well as additional and / or different types of blemishes on human skin, can be caused by various endogenous and / or exogenous factors.

[0003] Such spots may be similar in shape and size, and it may be difficult to determine which type of spot has formed on the skin or what the underlying cause of a particular spot is. This can lead to problems such as misidentifying such spots. Misidentification can then potentially lead to ineffective treatment for these spots. For example, a product designed to treat spots related to hemoglobin can be applied to spots related to melanin (or vice versa), and in some cases, it may not only be ineffective but potentially dangerous (such as applying prescription drugs to skin spots different from their original intended use). In particular, assuming the complexity of skin types, this problem worsens when considered across different users who may each be associated with different demographic attributes, races, and / or ethnicities. This causes problems in the diagnosis and treatment of various human skin conditions and characteristics. For example, prior art methods including personal consumer product testing can be time-consuming or error-prone (and in some cases ineffective). Additionally, users attempt to empirically experiment using various products or techniques, but without achieving satisfactory results and / or in some cases causing negative side effects that can affect the health or otherwise the visual appearance of the user's skin. Summary of the Invention Problems to be Solved by the Invention

[0004] For these reasons, there is a need for a digital imaging and artificial intelligence-based system and method for analyzing pixel data of a user's skin image and generating one or more user-specific skin spot classifications. Means for Solving the Problems

[0005] Generally, as described herein, a digital imaging and artificial intelligence-based system for analyzing pixel data of a user's skin area to generate one or more user-specific spot classifications is described. Such a digital imaging and artificial intelligence (AI)-based system provides a digital imaging and artificial intelligence-based solution to overcome the difficulties arising from identifying and treating various intrinsic and / or extrinsic factors or attributes that affect human skin health.

[0006] In the digital imaging and artificial intelligence-based systems described herein, a user can transmit specific user images to an imaging server (e.g., including one or more processors) or other computing device (e.g., on the user's mobile device). The imaging server or the user's computing device implements or runs an artificial intelligence-based skin learning model trained on pixel data from 10,000 (or more) images depicting the individual's skin or skin area. This skin learning model can generate at least one user-specific blemish classification, designed to address at least one identifiable feature within pixel data that includes at least a portion of the user's skin area, based on the user's skin image classification. For example, the user's skin image may include pixels or pixel data indicating blemishes (e.g., hemoglobin and / or melanin-related blemishes) or other attributes / conditions of a particular user's skin. In some embodiments, the user-specific blemish classification (and / or product-specific blemish classification) may be transmitted to the user's computing device via a computer network and rendered on a display screen. In other embodiments, user-specific images are not transmitted to the imaging server, and user-specific blemish classifications (and / or product-specific blemish classifications) are instead generated by a skin learning model that runs and / or is implemented locally on the user's mobile device and can be rendered on the mobile device's display screen by the mobile device's processor. In various embodiments, such renderings may include graphic representations, overlays, annotations, etc., to address features within the pixel data.

[0007] More specifically, a digital imaging and artificial intelligence-based system is disclosed, configured to analyze pixel data of a user's skin image to generate one or more user-specific skin spot classifications, as described herein. This digital imaging and artificial intelligence-based system is configured to analyze pixel data of a user's skin image to generate one or more user-specific skin spot classifications. This digital imaging and artificial intelligence-based system may comprise one or more processors and an imaging application (app) comprising computing instructions configured to run on one or more processors. This digital imaging and artificial intelligence-based system may further comprise a skin-based learning model accessible from the imaging app and trained on pixel data of multiple training images depicting each individual's skin. This skin-based learning model may be configured to output one or more spot classifications corresponding to one or more spot features in each individual's skin region. When the computing instructions of the imaging app are executed by one or more processors, they may cause one or more processors to receive a user's image, the image comprising a digital image such that it is captured by an imaging device, and the image comprising pixel data of at least a portion of the user's skin region. When the imaging application's computing instructions are executed by one or more processors, these processors can further analyze the images captured by the imaging device using a skin-based learning model to determine at least one blemish classification of the user's skin. The at least one blemish classification can be selected from one or more blemish classifications of the skin-based learning model. When the imaging application's computing instructions are executed by one or more processors, these processors can further generate user-specific skin recommendations designed to address at least one identifiable blemish feature within pixel data, including a portion of the user's skin area, based on the user's at least one blemish classification.

[0008] Furthermore, as described herein, a digital imaging and artificial intelligence-based method is disclosed for analyzing pixel data of a user's skin image to generate one or more user-specific skin spot classifications. This digital imaging and artificial intelligence-based method includes receiving a user image in one or more processors. This image includes a digital image captured by an imaging device and includes pixel data of at least a portion of the user's skin area. This digital imaging and artificial intelligence-based method may further include analyzing the image captured by the imaging device by a skin-based learning model running on one or more processors to determine at least one skin spot classification of the user's skin. The at least one skin spot classification can be selected from one or more skin spot classifications of the skin-based learning model. In various embodiments, the skin-based learning model may be trained using pixel data of multiple training images depicting each individual's skin, and the skin-based learning model is configured to output one or more skin spot classifications corresponding to one or more skin spot features of each individual's skin area. The digital imaging and artificial intelligence-based method may further include, by one or more processors, generating user-specific skin recommendations based on at least one blemish classification of the user's skin, which are designed to address at least one identifiable blemish feature within pixel data that includes a portion of the user's skin area.

[0009] Furthermore, a tangible, non-temporary, computer-readable medium is disclosed, as described herein, for storing instructions for analyzing pixel data of an image of the user's skin to generate one or more user-specific skin spot classifications. When executed by one or more processors, these instructions can cause one or more processors to receive an image of the user. The image may include a digital image captured by an imaging device and may include pixel data of at least a portion of the user's skin area. When executed by one or more processors, these instructions can further cause one or more processors to analyze the image captured by the imaging device using a skin-based learning model to determine at least one skin spot classification of the user. The at least one skin spot classification can be selected from one or more skin spot classifications of the skin-based learning model. In various embodiments, the skin-based learning model can be trained using pixel data of multiple training images depicting each individual's skin. The skin-based learning model may be configured to output one or more skin spot classifications corresponding to one or more skin spot features of each individual's skin area. When these instructions are executed by one or more processors, they can further cause one or more processors to generate user-specific skin recommendations designed to address at least one identifiable blemish feature within pixel data that includes a portion of the user's skin area, based on at least one blemish classification of the user's skin.

[0010] As disclosed above and in accordance with the disclosures herein, this disclosure includes improvements to the computer's capabilities or other technologies, since it discloses, for example, that an imaging server or other computing device (e.g., a user computer device) is enhanced by a trained (e.g., machine learning-trained) skin-based learning model to improve the intelligence or predictive capabilities of the server or computing device. A skin-based learning model running on an imaging server or computing device can more accurately identify one or more user-specific skin recommendations designed to address user-specific skin or blemish features, image classification of user skin areas, and / or at least one identifiable feature within pixel data including at least a portion of a user's skin area, based on pixel data of various individuals. In other words, this disclosure describes improvements to the computer's capabilities or “any other technology or technical field,” since the imaging server or user computing device is enhanced using multiple training images (e.g., 10,000 training images and associated pixel data as feature data) to accurately predict, detect, classify, or determine the pixel data of user-specific images, such as newly provided user images. This is superior to prior art because existing systems, at least, lack such predictive and classification capabilities and cannot accurately analyze user-specific images to output predictive results that address at least one identifiable feature within pixel data that includes at least a portion of a particular user's skin area.

[0011] For similar reasons, this disclosure describes or presents improvements to computing devices in the skincare and skincare products fields, and is also related to improvements to other technologies or technologies, as trained skin-based learning models running on imaging devices or computing devices improve fields such as skincare, chemical formulations, and / or skin classification and identification, and analyze user or individual images based on digital and / or artificial intelligence to output predictive results for user-specific pixel data to address at least one identifiable feature within pixel data that includes at least a portion of a particular user's skin area.

[0012] Furthermore, this disclosure describes or presents improvements to computing devices in the skincare and / or skincare products field, where trained skin-based learning models running on imaging devices or computing devices improve underlying computing devices (e.g., imaging servers and / or user computing devices), and such computing devices are also related to improvements to other technologies or technical fields as efficiency is improved by configuring, tuning, or adapting specific machine learning network architectures. For example, in some embodiments, fewer machine resources (e.g., processing cycles or memory storage) can be used by reducing the machine learning network architecture required to analyze images, including reducing depth, width, image size, or other machine learning-based dimensional requirements, thereby reducing computational resources. Such reductions free up computational resources in the underlying computing system, thereby making the system more efficient.

[0013] Furthermore, this disclosure describes or introduces improvements to computing devices in the field of security, and thus relates to improvements to other arts or fields, by describing or introducing improvements to computing devices in the field of security, where the user's image is preprocessed (e.g., cropped or otherwise modified) to define extracted or depicted skin areas of the user without depicting the user's personally identifiable information (PII). For example, a simple cropped or edited portion of the user's image may be used by the skin-based learning model described herein, which eliminates the need to transmit the user's private photographs over a computer network (such images may be susceptible to interception by third parties). Such features provide an improvement in security because the cropped or edited image, in particular, the image that may be transmitted over a network (e.g., the Internet), does not contain the user's PII information, i.e., the removal of PII (e.g., facial features) provides an improvement over conventional systems. Accordingly, the systems and methods described herein operate without requiring such non-essential information, which provides an improvement over conventional systems, e.g., an improvement in security. In addition, the use of cropped images, in at least some embodiments, allows the underlying system to store and / or process images with smaller data sizes, which in turn results in an overall increase in the performance of the underlying system because images with smaller data sizes require less storage memory and / or processing resources to be stored, processed, and / or otherwise manipulated by the underlying computer system.

[0014] In addition, the disclosure includes applying some of the claim elements together with, or by using, a specific machine, such as an imaging device, which is used to train a skin-based learning model and to capture images used to determine image classifications corresponding to one or more features of a user's scalp area.

[0015] Furthermore, this disclosure includes certain features that differ from well-known, conventional, or prior art activities, or unconventional processes that limit the claims to specific useful applications (for example, digital imaging and artificial intelligence-based systems and methods for analyzing pixel data of a user's skin image to generate one or more user-specific skin spot classifications).

[0016] The advantages will become more apparent to those skilled in the art from the following description of preferred embodiments illustrated and illustrated by examples. As will be understood, other and different embodiments are possible, and their details can be modified in various ways. Therefore, the drawings and description should be considered as illustrative and not as limiting. [Brief explanation of the drawing]

[0017] The figures described below illustrate various aspects of the systems and methods disclosed herein. Each figure depicts an embodiment of a particular aspect of the disclosed systems and methods, and it should be understood that each figure is intended to correspond to its possible embodiment. Furthermore, wherever possible, the following description refers to the reference figures included in the following figures, and features depicted in multiple figures are designated using consistent reference figures.

[0018] The drawings show the currently considered configuration, but it should be understood that this embodiment is not limited to the exact configuration and means shown. [Figure 1] This specification illustrates an example of a digital imaging and artificial intelligence-based system configured to analyze pixel data of a user's skin image to generate one or more user-specific skin spot classifications, according to various embodiments disclosed herein. [Figure 2] Examples of images and their associated pixel data that can be used to train and / or implement skin-based learning models are illustrated according to various embodiments disclosed herein. [Figure 3]This specification illustrates examples of digital imaging and artificial intelligence-based methods for analyzing pixel data of a user's skin image to generate one or more user-specific skin spot classifications, according to various embodiments disclosed herein. [Figure 4-1] Figures 4A to 4E illustrate exemplary images in which digital imaging and artificial intelligence-based algorithms are applied to analyze pixel data of a user's skin image to generate one or more user-specific skin spot classifications, according to various embodiments disclosed herein. [Figure 4-2] Figures 4A to 4E illustrate exemplary images in which digital imaging and artificial intelligence-based algorithms are applied to analyze pixel data of a user's skin image to generate one or more user-specific skin spot classifications, according to various embodiments disclosed herein. [Figure 5] This specification illustrates exemplary user interfaces rendered on the display screen of a user computing device according to various embodiments disclosed herein.

[0019] These figures illustrate preferred embodiments for illustrative purposes only. Alternative embodiments of the systems and methods illustrated herein may be adopted without departing from the principles of the invention described herein. [Modes for carrying out the invention]

[0020] The systems and methods described herein may be used to identify, classify, and / or treat skin blemishes or scars, which can be defined as contrasts caused by biological chromophores such as hemoglobin or melanin. These blemishes may include, but are not limited to, freckles, solar lentigines, melanosis, seborrheic keratosis, actinic keratosis, post-inflammatory hyperpigmentation, post-inflammatory erythema, moles, age spots, and darkened pores. The terms “scar” and “blemish” may be used interchangeably.

[0021] Skin blemishes and scars are difficult to remove, and there are few over-the-counter medications that are effective in restoring skin to its original, even state. It is hypothesized that blemishes and scars are formed through the following dynamic process: When the skin is damaged (e.g., acne including comedones, wounds, insect bites), local inflammation begins, and often one or more red spots are formed. The redness and inflammation are generally due to the chromophore hemoglobin. The amount of hemoglobin in the red spots may change over time, thereby altering the appearance of the spots. In some cases, an increase in melanin production may be observed in these red spots. This increase in melanin tends to darken the appearance of the spots. Since these spots may heal spontaneously over time, it is expected that the hemoglobin and / or melanin in these spots will decrease over time. However, in some cases, especially if the symptoms are severe, blemishes can remain on the skin for a long time. Compositions containing hydroxycinnamic acid (HCA) and niacinamide at a low pH have been found to reduce melanin and hemoglobin in persistent blemishes or scars. This disclosure describes systems and methods for identifying and / or classifying such stains, which enable effective treatment such as the selection and use of products and / or compositions.

[0022] Figure 1 illustrates an example of a digital imaging and artificial intelligence-based system configured to analyze pixel data of a user's skin image (e.g., one or more of images 202a, 202b, and / or 202c) according to various embodiments disclosed herein, in order to generate one or more user-specific skin spot classifications. Generally, as referred to herein, one or more spot classifications may include one or more hemoglobin type classifications (e.g., as shown herein for the exemplary image in Figure 4D) or melanin type classifications (e.g., as shown herein for the exemplary image in Figure 4E). However, it should be understood that additional and / or different types of skin spots can also be analyzed, identified, and / or classified according to the systems and methods described herein.

[0023] In the embodiment shown in Figure 1, the digital imaging and artificial intelligence-based system 100 may include a server 102 which may include one or more computer servers. In various embodiments, the server 102 may include multiple servers, and these servers may include multiple, redundant, or replicated servers as part of a server farm. In further embodiments, the server 102 may be implemented as a cloud-based server, such as a cloud-based computing platform. For example, the imaging server 102 may be one or more cloud-based platforms such as Microsoft Azure or Amazon AWS. The server 102 may include one or more processors 104 (i.e., CPUs) and one or more computer memories 106. In various embodiments, the server 102 may be referred to herein as the “imaging server”.

[0024] Memory 106 may include one or more forms of volatile and / or non-volatile fixed and / or removable memory, such as read-only memory (ROM), electronic programmable read-only memory (EPROM), random access memory (RAM), erasable electronic programmable read-only memory (EEPROM), and / or other hard drives, flash memory, microSD cards, etc. Memory 106 may store an operating system (OS) (e.g., Microsoft Windows, Linux, UNIX, etc.) that can facilitate functions, apps, methods, or other software, as considered herein. Memory 106 may also store a skin-based learning model 108. This learning model 108 may include an artificial intelligence-based model, such as a machine learning model, trained on various images (e.g., images 202a, 202b, and / or 202c), as described herein. Additionally, or alternatively, the skin-based learning model 108 may also be stored in a database 105 that is accessible to the imaging server 102 or otherwise communicated with. In addition, the memory 106 may also store machine-readable instructions and include one or more applications (e.g., imaging applications as described herein), one or more software components, and / or one or more application programming interfaces (APIs), which may be implemented to facilitate or perform any features, functions, or other disclosures described herein, such as any methods, processes, elements, or limitations, as illustrated, depicted, or explained in the various flowcharts, illustrative diagrams, schematics, figures, and / or other disclosures herein.For example, at least some of the applications, software components, or APIs may be image-based machine learning models or components, such as the skin-based learning model 108, and may include, or otherwise include, parts thereof, in which case each may be configured to facilitate their various functions considered herein. One or more other applications may be conceivable and should be understood to be executed by the processor 104.

[0025] The processor 104 may be connected to the memory 106 via a computer bus responsible for transmitting electronic data, data packets, or otherwise electronic signals between the processor 104 and the memory 106 in order to implement or perform machine-readable instructions, methods, processes, elements, or limitations, as illustrated, depicted, or described in the various flowcharts, illustrative diagrams, schematics, figures, and / or other disclosures herein.

[0026] The processor 104 may interface with memory 106 via a computer bus to execute an operating system (OS). The processor 104 may also interface with memory 106 via a computer bus to create, read, update, delete, or otherwise access or interact with data stored in memory 106 and / or database 104 (e.g., a relational database such as Oracle, DB2, MySQL, or a NoSQL-based database such as MongoDB). The data stored in memory 106 and / or database 105 may include all or part of the data or information described herein, for example, training images and / or user images (e.g., including one or more of images 202a, 202b, and / or 202c), zoomed, cropped, and / or segmentation-related images (e.g., 202azs, 202azs1, 202azs2, etc.), and / or other images and / or information of the user, including demographic attributes such as age, race, skin type, etc., or as otherwise described herein.

[0027] The imaging server 102 may further include communication components configured to communicate (e.g., transmit and receive) data to one or more networks or local terminals, such as the computer network 120 and / or terminal 109 (for rendering or visualization) described herein, via one or more external / network ports. In some embodiments, the imaging server 102 may include client-server platform technology such as ASP.NET, Java J2EE, Ruby on Rails, Node.js, web services, or online APIs, which are responsible for receiving and responding to electronic requests. The imaging server 102 may also implement client-server platform technology that can implement or perform machine-readable instructions, methods, processes, elements, or limitations, as illustrated, depicted, or described in the various flowcharts, illustrative diagrams, schematics, figures, and / or other disclosures herein, by interacting with memory 106 (including applications, components, APIs, data, etc., stored therein) and / or database 105 via a computer bus.

[0028] In various embodiments, the imaging server 102 may include, or interact with, one or more transceivers (e.g., WWAN, WLAN, and / or WPAN transceivers) that function in accordance with IEEE standards, 3GPP standards, or other standards and can be used to receive and transmit data via external / network ports connected to the computer network 120. In some embodiments, the computer network 120 may include a private network or a local area network (LAN). Additionally or alternatively, the computer network 120 may include a public network such as the Internet.

[0029] The imaging server 102 may further include or implement an operator interface configured to present information to an administrator or operator and / or receive input from an administrator or operator. As shown in Figure 1, the operator interface may provide a display screen (e.g., via terminal 109). The imaging server 102 may also provide I / O components (e.g., ports, capacitive or resistive touch-sensitive input panels, keys, buttons, lighting, LEDs), which may be directly accessible via the imaging server 102 or mounted on the server, or indirectly accessible via terminal 109 or mounted on the terminal. According to some embodiments, an administrator or operator may access the server 102 via terminal 109 to review and modify information, input training data or images, start training a skin-based learning model 108, and / or perform other functions.

[0030] As described herein, in some embodiments, the imaging server 102 may perform the functions considered herein as part of a “cloud” network, or otherwise communicate with other hardware or software components in the cloud to transmit, retrieve, or otherwise analyze the data or information described herein.

[0031] In general, computer programs or computer-based products, applications, or code (e.g., models such as AI models or other computing instructions described herein) may be stored on computer-usable storage media or tangible non-temporary computer-readable media (e.g., standard random-access memory (RAM), optical disks, universal serial bus (USB) drives, etc.) that have such computer-readable program code or computer instructions embodied within them, and the computer-readable program code or computer instructions may be installed on a processor 104 (e.g., working in relation to the respective operating system in memory 106) or otherwise adapted to be executed by that processor, so as illustrated, depicted, or described in the various flowcharts, illustrative diagrams, schematics, figures, and / or other disclosures herein, they may facilitate, implement, or execute machine-readable instructions, methods, processes, elements, or limitations. In this regard, program code can be implemented in any desired programming language, and can be implemented as machine code, assembly code, bytecode, interpreted source code, etc. (e.g., via Golang, Python, C, C++, C#, Objective-C, Java, Scala, ActionScript, JavaScript, HTML, CSS, XML, etc.).

[0032] As shown in Figure 1, the imaging server 102 is communicably connected to one or more user computing devices 111c1 to 111c3 via a computer network 120 and / or to 112c1 to 112c4 via base stations 111b and 112b. In some embodiments, base stations 111b and 112b may include cellular base stations such as cell towers, which communicate with one or more user computing devices 111c1 to 111c3 and 112c1 to 112c4 via wireless communication 121 based on one or more of the various mobile phone standards, including NMT, GSM, CDMA, UMMTS, LTE, 5G, etc. Additionally, or alternatively, base stations 111b and 112b may include, by non-limiting examples, routers, wireless switches, or other such wireless connection points that communicate with one or more user computing devices 111c1-111c3 and 112c1-112c4 via wireless communication 122 based on one or more of various wireless standards, including IEEE 802.11a / b / c / g (WIFI), Bluetooth standards, etc.

[0033] One or more user computing devices 111c1 to 111c3 and / or 112c1 to 112c4 may include a mobile device and / or client device for accessing and / or communicating with the imaging server 102. Such a mobile device may comprise one or more mobile processors and / or imaging devices for capturing images such as images described herein (e.g., one or more of images 202a, 202b, and / or 202c). In various embodiments, user computing devices 111c1 to 111c3 and / or 112c1 to 112c3 may, by non-limiting examples, include mobile phones (e.g., cellular phones), tablet devices, personal data assistance (PDAs), etc., including Apple iPhone or iPad devices, or Google Android-based mobile phones or tablets.

[0034] In additional embodiments, the user computing device 112c4 may be a portable microscope device, such as a skin mirror, that the user can use to capture detailed images of the user's skin. Specifically, the portable microscope device 112c4 may include a microscope camera configured to capture images (e.g., one or more of images 202a, 202b, and / or 202c) at a nearly microscopic level of skin areas of the user's skin. Unlike, for example, any of the user computing devices 111c1-111c3 and 112c1-112c3, the portable microscope device 112c4 may capture detailed high-magnification (e.g., 2-megapixel images at 60-200x magnification) images of the user's skin while maintaining physical contact with the user's skin. As a particular example, the portable microscope device 112c4 may be the API 100 SKIN ANALYSIS device developed by NERA SOLUTIONS LTD. In certain embodiments, the portable microscope device 112c4 may also include a display or user interface configured to show the captured images and / or the results of image analysis to the user.

[0035] Additionally, or alternatively, the portable microscope device 112c4 may be communicatively coupled to a user computing device 112c1 (e.g., the user's mobile phone) via a WIFI connection, a Bluetooth connection, and / or any other suitable wireless connection, and the portable microscope device 112c4 may be compatible with various operating platforms (e.g., Windows, iOS, Android, etc.). Thus, the portable microscope device 112c4 may transmit captured images to the user computing device 112c1 for analysis and / or display to the user. Furthermore, the portable microscope device 112c4 may be configured to capture high-quality video of the user's skin and stream the high-quality video of the user's skin to the display of the portable microscope device 112c4 and / or the communicatively coupled user computing device 112c1 (e.g., the user's mobile phone). In certain additional embodiments, each component of the portable microscope device 112c4 and the communicatively connected user computing device 112c1 may be incorporated into a single device.

[0036] In additional embodiments, user computing devices 111c1-111c3 and / or 112c1-112c3 may include retail computing devices. The retail computing devices may comprise user computer devices configured in the same or similar manner as, for example, the mobile devices described herein with respect to user computing devices 111c1-111c3, including having a processor and memory for implementing a skin-based learning model 108 as described herein, or for communicating with the skin-based learning model 108 (e.g., via a server 102). Additionally or alternatively, the retail computing devices may be located, installed, or otherwise positioned within a retail environment to enable users and / or customers of the retail environment to utilize digital imaging and artificial intelligence-based systems and methods on-site within the retail environment. For example, the retail computing devices may be installed within a kiosk for user access. Users may then upload or transfer images (e.g., from a user mobile device) to the kiosk in order to implement the digital imaging and artificial intelligence-based systems and methods described herein. Additionally, or alternatively, the kiosk may be configured with a camera to allow the user to take their new images (e.g., in a private manner, where guaranteed) for upload and transfer. In such embodiments, the user or consumer themselves may use the retail computing device to receive and / or render user-specific electronic stain classifications on the display screen of the retail computing device, as described herein.

[0037] Additionally, or alternatively, the retail computing device may be a mobile device (as described herein) carried by an employee or other personnel in the retail environment to interact with a user or consumer on-site. In such embodiments, the user or consumer may interact with an employee or other personnel in the retail environment via the retail computing device (for example, by transferring images from the user's mobile device to the retail computing device, or by capturing new images with the camera of the retail computing device) to receive and / or render a user-specific electronic skin classification on the display screen of the retail computing device, as described herein.

[0038] In various embodiments, one or more user computing devices 111c1-111c3 and / or 112c1-112c4 may implement or run an operating system (OS) or mobile platform such as Apple's iOS and / or Google's Android operating system. Any of the one or more user computing devices 111c1-111c3 and / or 112c1-112c4 may include one or more processors and / or one or more memories for storing, implementing, or running computing instructions or code, such as a mobile application or a home or personal assistant application, as described in various embodiments herein. As shown in Figure 1, the skin-based learning model 108a and / or the imaging application described herein, or at least a part thereof, may be stored locally in the memory of a user computing device (e.g., user computing device 111c1). In some embodiments, the skin-based learning model 108a installed on the computing device may comprise the same skin-based learning model 108 installed on the server 102. Additionally, or alternatively, the skin-based learning model 108a may include a portion of the skin-based learning model 108 installed on the server 102. The skin-based learning model may, in some embodiments, be fully installed on the user's computing device, fully installed on the server 102, or partially installed on both the user's computing device and the server 102, in which case communication between the skin-based learning model 108a and the skin-based learning model 108 should be understood to occur via the computer network 120. Generally, when the skin-based learning model is referred to herein, it refers to either or both of the skin-based learning model 108 and / or the skin-based learning model 108a.

[0039] User computing devices 111c1 to 111c3 and / or 112c1 to 112c4 may include wireless transceivers for receiving and transmitting wireless communications 121 and / or 122 to and from base stations 111b and / or 112b. In various embodiments, pixel-based images (e.g., images 202a, 202b, and / or 202c) may be transmitted to an imaging server 102 via a computer network 120 for training a model (e.g., a skin-based learning model 108) and / or for image analysis as described herein.

[0040] In addition, one or more user computing devices 111c1-111c3 and / or 112c1-112c4 may include imaging devices and / or digital video cameras for capturing or photographing digital images and / or frames (which may be one or more of images 202a, 202b, and / or 202c). Each digital image may include pixel data for training or implementing a model such as an AI or machine learning model, as described herein. For example, any imaging device and / or digital video camera among the user computing devices 111c1-111c3 and / or 112c1-112c4 may be configured to photograph, capture, or otherwise generate digital images (e.g., pixel-based images 202a, 202b, and / or 202c), and in at least some embodiments, such images may be stored in the memory of the respective user computing device. Additionally or alternatively, such digital images may also be transmitted to and / or stored in the memory 106 and / or database 105 of the server 102.

[0041] Furthermore, each of the one or more user computer devices 111c1-111c3 and / or 112c1-112c4 may include a display screen for displaying graphics, images, text, blemish classifications, skin products, data, pixels, features, and / or other such visualizations or information as described herein. In various embodiments, graphics, images, text, blemish classifications, skin products, data, pixels, features, and / or other such visualizations or information may be received from the imaging server 102 for display on one or more of the display screens of the user computer devices 111c1-111c3 and / or 112c1-112c4. Additionally or alternatively, a user computing device, such as the one described herein with respect to Figure 5, may, at least in part, have, implement, have access to, render, or otherwise expose an interface or guided user interface (GUI) for displaying text and / or images on its display screen.

[0042] In some embodiments, computing instructions and / or applications running on a server (e.g., server 102) and / or a mobile device (e.g., mobile device 111c1) may be communicatively connected to analyze pixel data of a user's skin image to generate one or more user-specific skin spot classifications, as described herein. For example, one or more processors of server 102 (e.g., processor 104) may be communicatively coupled to a mobile device via a computer network (e.g., computer network 120). In such embodiments, the imaging application may include a server application portion configured to run on one or more processors of a server (e.g., server 102) and a mobile application portion configured to run on one or more processors of a mobile device (e.g., one or more user computing devices 111c1-111c3 and / or any of 112c1-112c3) and / or a standalone imaging device (e.g., user computing device 112c4). In such embodiments, the server application portion is configured to communicate with the mobile application portion. The server application portion or the mobile application portion may be configured to implement or partially implement one or more of the following: (1) receiving images captured by the imaging device; (2) determining at least one spot classification for the user's skin area; (3) generating user-specific spot classifications; and / or (4) sending user-specific recommendations to the application portion of the computing device.

[0043] Figure 2 illustrates an exemplary image 202az and its associated pixel data that may be used to train and / or implement a skin-based learning model according to various embodiments disclosed herein. In various embodiments, as shown in Figure 2, image 202az may be an image captured by a user. In this embodiment, image 202az represents and is depicted as a zoomed or cropped version of image 202a in Figure 1. Image 202az (as well as images 202a, 202b, and / or 202c) may be transmitted to server 102 via computer network 120, as shown in Figure 1. It should be understood that such images may be captured by the user themselves, or additionally or alternatively, by others such as retailers who use and / or transmit such images on behalf of the user.

[0044] More generally, digital images such as exemplary images 202a, 202b, and 202c may be collected or aggregated by the imaging server 102 and may be analyzed by and / or used to train a skin-based learning model (e.g., an AI model such as a machine learning imaging model as described herein). Each of these images may contain pixel data within each image that includes feature data corresponding to each user's skin region. The pixel data may be captured by one imaging device among the user computing devices (e.g., one or more user computer devices 111c1 to 111c3 and / or 112c1 to 112c4).

[0045] Regarding digital images as described in this specification, pixel data (e.g., pixel data 202ap in FIG. 2) includes individual points or squares of data within the image, and each point or square represents a single pixel (e.g., each of pixel 202ap1, pixel 202ap2, and pixel 202ap3) within the image. Each pixel can be at a specific location within the image. Additionally, each pixel can have a specific color (or the absence thereof). The color of a pixel can be determined by the color format associated with the specific pixel and the associated channel data. For example, a common color format is the 1976 CIELAB (referred to herein as the "CIE L * -a * -b * " or simply the "L * a * b * " color format) color format, which is configured to mimic human color perception. That is, the L * a * b * color format is designed such that the amount of numerical change in the three values (e.g., L * a * b * ) representing the L * a * b * color format is approximately the same as the amount of change visually perceived by humans. This color format is advantageous, for example, because the L * a * b * color gamut (e.g., the complete subset of colors included as part of the color format) includes both the color gamut of the red (R), green (G), and blue (B) (collectively RGB) color format and the color gamut of the cyan (C), magenta (M), yellow (Y), and black (K) (collectively CMYK) color format.

[0046] In the L * a * b * color format, color is seen as a point in three-dimensional space as defined by the three-dimensional coordinate system (L * a * b * ), and the L * data, a* data, and b * Each piece of data can correspond to an individual color channel and can therefore be referred to as channel data. In this three-dimensional coordinate system, L * The axis describes the brightness (luminance) of colors, with values ​​ranging from 0 (black) to 100 (white). * The axis represents positive a, indicating the red hue. * Value (+a * ) and negative a which shows a green hue * Value (-a * ) Explain the ratio of green or red in the color. * The axis represents positive b, indicating a yellow hue. * Value (+b * ) and negative b which indicates a blue hue * Value (-b * ) Describe the ratio of blue or yellow in a color. Generally, a * axis and b * The value corresponding to the axis is a * axis and b * The axis may be unlimited so that it can include any suitable numerical value to represent the axis boundary. However, a * axis and b * The axis typically includes lower and upper limits in the range of approximately 150 to -150. Thus, each pixel color value is L * a * , and b * Represented as a 3-tuple of values, it can create the final color of a particular pixel.

[0047] As another example, common color formats include the Red-Green-Blue (RGB) format, which has red, green, and blue channels. That is, in the RGB format, pixel data is represented by three numerical RGB components (red, green, and blue), which can be called channel data, to manipulate the color of the area of ​​the pixel in the image. In some implementations, the three RGB components can be represented as three 8-bit numbers per pixel. Using three 8-bit bytes (one byte for each of RGB), a 24-bit color can be produced. Each 8-bit RGB component can have 256 possible values ​​in the range of 0 to 255 (i.e., in a radix 2 binary system, an 8-bit byte can contain one of 256 numerical values ​​in the range of 0 to 255). This channel data (R, G, and B) can be assigned values ​​from 0 to 255 that can be used to set the color of a pixel. For example, three values ​​such as (250, 165, 0), meaning (red=250, green=165, blue=0), can represent one orange pixel. As a further example, (red=255, green=255, blue=0) means red and green are fully saturated (255 is the brightness possible with 8 bits), blue is absent (zero), and the resulting color is yellow. As yet another example, black has the RGB values ​​(red=0, green=0, blue=0), and white has the RGB values ​​(red=255, green=255, blue=255). Gray has the characteristic of having equal or similar RGB values; for example, (red=220, green=220, blue=220) is light gray (close to white), and (red=40, green=40, blue=40) is dark gray (close to black).

[0048] In this way, the combination of the three RGB values ​​creates the final color for a particular pixel. In a 24-bit RGB color image, using 3 bytes to define color, there can be 256 levels of red, 256 levels of green, and 256 levels of blue. This provides 256 × 256 × 256, or 16.7 million possible combinations or colors for a 24-bit RGB color image. Therefore, the RGB data value of a pixel indicates the degree of color or light, consisting of each of the red, green, and blue pixels. The three colors and their intensity levels combine at that image pixel, i.e., at that pixel location on the display screen, to illuminate the display screen with that color at that location. However, it should be understood that other bit sizes, such as fewer or more bits, e.g., 10 bits, can be used to result in fewer or more overall colors and ranges.

[0049] Overall, various pixels positioned together within a grid pattern (e.g., pixel data 202ap) form a digital image or a part thereof. A single digital image can contain thousands or millions of pixels. Images can be captured, generated, stored, and / or transmitted in several formats, such as JPEG, TIFF, PNG, and GIF. These formats use pixels to store or represent images.

[0050] Referring to Figure 2, the example image 202az illustrates a user or individual's skin area. More specifically, image 202az includes pixel data, which includes pixel data 202ap defining the user or individual's skin area. Pixel data 202ap includes multiple pixels, including pixel 202ap1, pixel 202ap2, and pixel 202ap3. In the example of Figure 202a, pixels 202ap1, 202ap2, and 202ap3 each represent skin features corresponding to the image classification of the skin area. Generally, in various embodiments, the features of the user's skin or skin area may include one or more hemoglobin-related spots and / or melanin-related spots. Each of these features may be determined from or based on one or more pixels in the digital image (e.g., image 202az). For example, with respect to image 202az, pixels 202ap1 and 202ap2 each represent relatively bright pixels (e.g., relatively high L) located within the pixel data 202ap in the skin area of ​​the user's skin. * Pixels with a value (and / or relatively yellow pixels (e.g., relatively high or positive b)) and / or relatively yellow pixels (e.g., pixels with a value of b) * Pixel 202ap3 is a pixel with a value, which may represent the normal or more common value for the user's skin. However, pixel 202ap3 is a darker pixel (e.g., negative or low relative L). * Pixels with a value) and / or redder pixels (e.g., positive or higher relative a) *Pixel data 202ap may include pixels with values, each of which may represent a blemish associated with melanin or hemoglobin at that location in the image of the user's skin. Such pixel features may be used to train a skin-based learning model (e.g., skin-based learning model 108) to generate user-specific skin recommendations designed to address at least one identifiable blemish feature within the pixel data, which includes a portion of the user's skin area. In addition to pixels 202ap1, 202ap2, and 202ap3, pixel data 202ap may include various other pixels, e.g., various other skin areas and / or parts of skin, which may be used for analysis and / or model training and / or analysis using a pre-trained model such as skin-based learning model 108 described herein. For example, pixel data 202ap may further include pixels representing blemish features, and in various embodiments, in addition to the color of the blemish, grouping of such pixels at a particular location in the image (similar L * a * b * Pixels having values ​​and / or RGB values ​​provide training information for stain classification as described herein.

[0051] Digital images such as training images, user-submitted images, or otherwise digital images (e.g., any of images 202a, 202b, and / or 202c) may be or may contain cropped images. Generally, a cropped image is an image from which one or more pixels have been removed, deleted, or hidden from the originally captured image. In some embodiments, each of one or more of the multiple training images (e.g., any of images 202a, 202b, and / or 202c), or the user's image, contains at least one cropped image depicting a skin region with only one blemish feature. For example, referring to Figure 2, image 202az represents at least a portion of the original image. Cropped portion 202ac1 represents the first portion cropped from image 202az, removing the portion of the user's skin that may not contain an easily identifiable blemish feature (outside of cropped portion 202ac1). As a further example, the cropped portion 202ac2 represents a second cropped portion of image 202az, which removes a portion of the image (outside the cropped portion 202ac2) that may not contain blemish features that are as easily identifiable as the features contained in the cropped portion 202ac2 and therefore may not be very useful as training data. In various embodiments, the analysis and / or use of cropped images for training results in improved accuracy of skin-based learning models. It also improves the efficiency and performance of the underlying computer system in that such a system processes, stores, and / or transmits smaller-sized digital images. Furthermore, images may be transmitted that are cropped or have extracted or depicted the user's skin area without depicting the user's personally identifiable information (PII). Such cropped images offer improved security, namely, the removal of PII provides an improvement over conventional systems, because cropped or edited images, in particular images that may be transmitted over a network (e.g., the Internet), are more secure because they do not contain the user's PII information.Importantly, the systems and methods described herein can operate without requiring such non-essential information, which provides improvements over conventional systems, such as enhanced security and performance. Furthermore, while Figure 2 may depict and represent a cropped image, it should be understood that other image types, including but not limited to the original uncropped image (e.g., original image 202a) and / or cropped images of other types / sizes (e.g., cropped portion 202ac1 of image 202az), may be used or substituted in the same way.

[0052] The disclosure relating to image 202az in Figure 2 applies in the same or similar manner to other digital images described herein, including, for example, images 202a, 202b, and / or 202c, and it should be understood that such images also include pixels that may be analyzed and / or used for training models, as described herein.

[0053] Furthermore, as described herein, digital images of a user's skin can depict a variety of skin features, and these features can be used to train skin-based learning models among diverse users with varying skin features. For example, as illustrated in images 202a, 202b, and 202c, these user skin regions include skin features (e.g., blemishes) of the user's skin region, which are identifiable by the pixel data of each image. These skin features may include, for example, features indicating hemoglobin and / or melanin, and may include individual skin regions or features (e.g., blemishes) present in one or more locations distributed across the user's skin.

[0054] In various embodiments, the digital images (e.g., images 202a, 202b, and 202c) may include multiple angles or viewpoints depicting each individual or user's skin, whether used as training images depicting an individual or as images depicting a user or individual for analysis and / or blemish classification. That is, each image among one or more training images, or an image of a user, may include multiple angles or viewpoints depicting each individual's or user's skin area. The multiple angles or viewpoints may include different views, positions, proximity of the user and / or background, lighting conditions, or the environment in which the user is otherwise positioned in a particular image. For example, Figure 1 includes skin images (e.g., 202a, 202b, and 202c) depicting each individual's and / or user's skin area, captured at different angles and using different lighting conditions (e.g., visible, UV). Such images can be used for training, analyzing, and / or classifying user-specific blemishes in skin-based learning models, as described herein.

[0055] Figure 3 illustrates exemplary digital imaging and artificial intelligence-based methods 300, 300 for analyzing pixel data of a user's skin image to generate one or more user-specific skin spot classifications, according to various embodiments disclosed herein. In block 302, method 300 may include receiving a user image in one or more processors. This image may include a digital image (e.g., any of images 202a / 202az, 202b, and / or 202c) captured by an imaging device (e.g., an imaging device of mobile device 111c1). Furthermore, this image may include pixel data of at least a portion of the user's skin area. In various embodiments, one or more processors may include a processor 104 of server 102. Additionally or alternatively, one or more processors may include a processor of a mobile device (e.g., computing device 111c1). The images used in method 300, and more generally the images described herein, are pixel-based images captured by an imaging device (e.g., an imaging device of user computing device 111c1). In some embodiments, the image may include or refer to a plurality of images, such as a plurality of images (e.g., frames) collected using a digital video camera. A frame may include a sequence of images that define motion, and may include motion, video, etc.

[0056] In block 304, method 300 further includes analyzing images captured by an imaging device using a skin-based learning model (e.g., skin-based learning model 108) running on one or more processors to determine at least one blemish classification of the user's skin. In the example in Figure 3, at least one blemish classification is selected from one or more blemish classifications of the skin-based learning model (e.g., skin-based learning model 108). In various embodiments, the skin-based learning model (e.g., skin-based learning model 108) is trained with pixel data from multiple training images (e.g., any of images 202a / 202az, 202b, and / or 202c) depicting the skin of each individual. Once trained, the skin-based learning model is configured to output one or more blemish classifications corresponding to one or more blemish features (e.g., hemoglobin or melanin) in each individual's skin region.

[0057] In various embodiments, a skin-based learning model (e.g., skin-based learning model 108) is an AI-based model trained with at least one artificial intelligence (AI) algorithm. Training the skin-based learning model 108 involves performing image analysis on training images to construct the weights of the skin-based learning model 108 and its underlying algorithm (e.g., machine learning or artificial intelligence algorithm) used to predict and / or classify future images. For example, in various embodiments of this specification, the generation of the skin-based learning model 108 involves training the skin-based learning model 108 using multiple training images of multiple individuals (e.g., images 202a, 202b, 202c). Each of these training images contains pixel data, each depicting a skin region or other skin of an individual. In some embodiments, one or more processors of a server or cloud-based computing platform (e.g., imaging server 102) may receive multiple training images of multiple individuals via a computer network (e.g., computer network 120). In such embodiments, the server and / or cloud-based computing platform can train the skin-based learning model using the pixel data of the multiple training images. Furthermore, in some embodiments, the skin-based learning model may be further trained using the user's demographic data (e.g., data indicating race, skin color, etc.) and each user's environmental data (e.g., sunlight, geography, weather conditions, etc.). In such embodiments, the blemish classification generated by the skin-based learning model may be further based on the user's demographic and environmental data provided by the specific user.

[0058] In various embodiments, the machine learning image models described herein (e.g., skin-based learning model 108) may be trained using supervised or unsupervised machine learning programs or algorithms. These machine learning programs or algorithms may utilize neural networks, which may be convolutional neural networks, vision transformers, deep learning neural networks, or composite learning modules or programs that learn two or more features or feature datasets (e.g., pixel data) in a particular domain of interest. These machine learning programs or algorithms may also include natural language processing, semantic analysis, automated reasoning, regression analysis, support vector machine (SVM) analysis, decision tree analysis, random forest analysis, K-nearest neighbor analysis, Naive Bayes analysis, clustering, reinforcement learning, and / or other machine learning algorithms and / or techniques. In some embodiments, artificial intelligence and / or machine learning-based algorithms may be included as libraries or packages that run on the imaging server 102. For example, libraries may include TENSORFLOW-based libraries, PYTORCH libraries, and / or SCIKIT-LEARN Python libraries.

[0059] Machine learning may include identifying and recognizing patterns in existing data (e.g., identifying skin features in image pixel data, such as blemishes, color, or discoloration, as described herein), thereby facilitating prediction or identification of subsequent data (e.g., using a model on new pixel data of a new image to determine or generate a user-specific blemish classification designed to address at least one identifiable feature within the pixel data that includes at least a portion of the user's skin area).

[0060] Machine learning models, such as the skin-based learning models described herein in some embodiments, are created and trained on exemplary data (e.g., “training data” and associated pixel data) inputs or data (which may be called “features” and “labels”) to make valid and reliable predictions about new inputs, such as test-level or production-level data or inputs. In supervised machine learning, a machine learning program running on a server, computing device, or otherwise on a processor is provided with exemplary inputs (e.g., “features”) and their associated or observed outputs (e.g., “labels”), and the machine learning program or algorithm may determine or discover rules, relationships, patterns, or otherwise machine learning “models” that map such inputs (e.g., “features”) to outputs (e.g., labels) by, for example, determining and / or assigning weights or other metrics to its model across various feature categories of the model. Such rules, relationships, or otherwise models are then provided with subsequent inputs so that the model can run on a server, computing device, or otherwise on a processor and predict expected outputs based on the discovered rules, relationships, or models.

[0061] In unsupervised machine learning, a server, computing device, or processor may need to find its own structure in unlabeled, exemplary inputs, in which case, for example, multiple training iterations are performed by the server, computing device, or processor to train multiple generations of models until a satisfactory model is produced, e.g., a model that provides sufficient predictive accuracy when given test-level or production-level data or inputs.

[0062] Supervised learning and / or unsupervised machine learning may also involve retraining, retraining, or otherwise updating a model with new or different information, which may include information received, ingested, generated, or otherwise used over time. Disclosures herein may utilize one or both of such supervised or unsupervised machine learning techniques.

[0063] In various embodiments, a skin-based learning model (e.g., skin-based learning model 108) can be trained by one or more processors (e.g., one or more processors 104 of server 102, and / or processors of computer user devices such as mobile devices) using pixel data from multiple training images (e.g., any of images 202a, 202b, and / or 202c). In various embodiments, the skin-based learning model (e.g., skin-based learning model 108) is configured or trained to output one or more features of a user's skin or skin region for a particular image. In these embodiments, one or more features of the skin or skin region may differ based on the demographic attributes and / or ethnicity of one or more individual users shown in each training image, such as features typically associated with different races, genomes, and / or geographic locations, or naturally occurring features related to these demographic attributes and / or ethnicity. Furthermore, a skin-based learning model (e.g., skin-based learning model 108) can generate user-specific blemish classifications for each individual represented in each training image, based on each individual's ethnic and / or demographic attribute values.

[0064] In various embodiments, image analysis may include training a machine learning-based model (e.g., skin-based learning model 108) based on pixel data of images depicting each individual's skin or skin area. Additionally or alternatively, image analysis may use a previously trained machine learning imaging model to train pixel data (e.g., L * a * , and b *This may include determining the classification of one or more images of an individual, a user's skin, or a skin area based on the L values ​​(including and / or RGB values). For example, the model weights could be the various L values ​​of individual pixels in a given image. * a * b * It can be trained by analyzing the values. For example, dark or low L * Value (for example, L) * Pixels with a value less than 50 may indicate image regions where hemoglobin and / or melanin are present. Similarly, slightly brighter L * Value (for example, L) * Pixels with a value greater than 50 may indicate the absence of melanin or hemoglobin. Furthermore, high / low a * The value may indicate areas of skin containing more / less melanin and / or hemoglobin. L indicates skin tone. * a * b * If pixels with values ​​are located within a particular image, or surrounded by groups or sets of pixels with melanin or hemoglobin-derived color tones, a skin-based learning model (e.g., skin-based learning model 108) can determine the image classification or blemish classification of a user's skin regions and associated blemishes identified within a particular image. In this way, a machine learning image model can be trained or used with pixel data from 10,000 training images (e.g., data showing details of each individual's skin region) to determine the image classification of a user's skin regions and their various blemish classifications.

[0065] Referring further to Figure 3, in block 306, Method 300 further includes generating user-specific skin recommendations by one or more processors, which are designed to address at least one identifiable blemish feature in pixel data including a portion of the user's skin area, based on at least one blemish classification of the user's skin. In various embodiments, when computing instructions of the imaging app are executed by one or more processors, one or more processors can generate skin type codes determined based on user-specific blemish classifications, which are designed to address at least one identifiable blemish feature in pixel data including a portion of the user's skin area. In some embodiments, the skin type codes may comprise IDs within a range or class of IDs for a particular skin type classification. The IDs or codes may also be arranged in terms of the degree of severity for a particular type, for example, lower values ​​indicating a lower severity of melanin and / or hemoglobin on the skin, and vice versa. These skin type codes are L * a * b * The values ​​may include RGB values, the percentage of melanin and / or hemoglobin detected in the skin, and / or any other appropriate codes. For example, in these embodiments, the user-specific spot classification may include mean / sum values ​​corresponding to each ID / value associated with each feature / attribute analyzed as part of the skin-based learning model (e.g., skin-based learning model 108), such as the hemoglobin to melanin ratio, which is detected, identified, or classified by the skin-based learning model 108. To illustrate, if a user has a high skin quality code (e.g., a classification of a value of 7 for hemoglobin and 17 for melanin), the user may receive a user-specific spot classification of "poor" or "unhealthy". In contrast, if a user receives a low score (e.g., 1 or 11, respectively), the user may receive a user-specific spot classification of "good" or "healthy".

[0066] In yet another embodiment, computing instructions in the imaging application may cause one or more processors to further record an image of the user captured by the imaging device at an initial time and track changes in the user's skin area over time. The second image may include pixel data of at least a portion of the user's skin area. Computing instructions may also cause one or more processors to analyze the second image captured by the imaging device using a skin-based learning model to determine a second image classification of the user's skin area selected from one or more image classifications of the skin-based learning model at a second time, and to generate a new user-specific blemish classification relating to at least one blemish feature, or its absence, that is identifiable within the pixel data of the second image, which includes at least a portion of the user's skin area, based on a comparison between the image of the user's skin area and the second image and / or image classification and the second image classification.

[0067] In an additional embodiment, at least one user-specific skin recommendation is displayed on the display screen of the computing device along with instructions for treating at least one blemish feature identifiable within pixel data that includes a portion of the user's skin area. In some embodiments, the user-specific skin recommendation may be rendered on the display screen in real time or near real time, during or after reception of the user's image. For example, instructions may provide the user with information to reduce or eliminate hypermelanin production in skin areas identifiable in the image (e.g., avoid direct sunlight exposure). This embodiment is further illustrated herein by Figure 5.

[0068] In further embodiments, at least one user-specific spot recommendation includes a product recommendation relating to a manufactured product. The manufactured product may include a pharmaceutical, therapeutic, or other product for treating at least one identifiable spot feature in the pixel data. For example, the product may include a composition such as a cream, in which low-pH hydroxycinnamic acid (HCA) and niacinamide can reduce melanin and hemoglobin in persistent spots or scars. In some embodiments, a product-based user-specific skin recommendation may be displayed on the display screen of a computing device along with instructions for treating at least one identifiable spot feature in the pixel data, which includes a portion of the user's skin area, using the manufactured product. Furthermore, in some embodiments, the computing instructions may further cause one or more processors to initiate a manufactured product for shipment to the user based on at least one user-specific skin recommendation.

[0069] With regard to the recommendation of manufactured products, in some embodiments, one or more processors (e.g., an imaging server 102 and / or a user computing device such as a user computing device 111c1) may generate a modified image, for example, based on at least one originally received image of the user. In such embodiments, the modified image may depict a rendering of what the user's skin or area of ​​skin is expected to look like after treating at least one feature with the manufactured product. For example, the modified image may be modified by updating, smoothing, or changing the color of pixels in the image to represent possible or expected changes after treating at least one feature in the pixel data with the manufactured product. The modified image may then be rendered on the display screen of a user computing device (e.g., user computing device 111c1).

[0070] In various embodiments, user-specific skin recommendations may be generated by the user's computing device (e.g., user computing device 111c1) and / or a server (e.g., pixel server 102). For example, in some embodiments, as described herein with respect to Figure 1, the pixel server 102 can analyze a user image remotely from the user computing device and determine the user's skin image classification, a user-specific blemish classification designed to address at least one identifiable feature within the pixel data containing at least a portion of the user's skin area, and / or the user-specific skin recommendation itself. For example, in such embodiments, the pixel server or cloud-based computing platform (e.g., pixel server 102) receives at least one image containing pixel data of at least a portion of the user's skin area via the computer network 120. The server or cloud-based computing platform can then run a skin-based learning model (e.g., skin-based learning model 108) and generate user-specific skin recommendations based on the output of the skin-based learning model (e.g., skin-based learning model 108). The server or cloud-based computing platform may then transmit user-specific skin recommendations to the user computing device via a computer network (e.g., computer network 120) for rendering on the user computing device's display screen. For example, in various embodiments, at least one user-specific skin recommendation may be rendered in real time or near real time on the user computing device's display screen while or after receiving an image containing skin areas of the user's skin.

[0071] Figures 4A to 4E illustrate exemplary images (e.g., Image 202a) to which a digital imaging and artificial intelligence-based algorithm 400 is applied to analyze pixel data of a user's skin image to generate one or more user-specific skin spot classifications, according to various embodiments disclosed herein. In some embodiments, the algorithms of Figures 4A to 4E may be implemented as or as part of a method 300 described herein.

[0072] As shown in Figure 4A, an image is input for processing by algorithm 400 or selected by other means. In this example, the image in Figure 4A is image 202a. In step 402, image calibration may be applied. More specifically, image calibration may include a preprocessing step that includes applying a color adaptation algorithm and / or a white balance (e.g., deep learning white balance) algorithm to an image (e.g., image 202a) to prepare the image for input to a skin-based learning model (e.g., skin-based learning model 108). Such preprocessing can improve the accuracy of the skin-based learning model (e.g., skin-based learning model 108) by removing data noise or other irrelevant elements. For example, image calibration including color adaptation may include taking a negative or other deformation of the original image to remove classification anomalies caused by differences in skin color. As a further example, image calibration may include white balance, where pixel values ​​are enhanced, reduced, or otherwise modified from one image to another to remove classification anomalies caused by differences in skin color.

[0073] The image calibration algorithm can be applied to both the input images for classification and the training images used to train the model. For example, each of several training images (e.g., any one or more of images 202a, 202b, and / or 202c, etc.) has an image calibration applied to modify the image for training purposes to enhance spot classification. The image calibration algorithm can then be applied to subsequent input images of the user prior to analyzing the user's images using a skin-based learning model, so that all images used for training and subsequent inputs have the same image calibration algorithm applied to them.

[0074] In some embodiments, image calibration may also involve cropping or zooming the original image to remove foreign features and thus reduce the file size. For example, as shown in Figure 4B, image 202az is a cropped or zoomed variation of image 202a, and image 202az was obtained by applying image calibration in step 402.

[0075] In step 404, the preprocessed image (e.g., image 202az) can be input to the skin-based learning model 108. In the examples of Figures 4A to 4E, the skin-based learning model 108 is an ensemble-based AI model, such as a transformer model, which includes (i) a segmentation model configured to generate segmentation mappings of one or more blemishes in the skin region of the image, and (ii) a prediction or classification model configured to analyze the pixel data of the segmentation mappings of one or more blemishes. In one example, the segmentation model may comprise a UNET-based segmentation model configured to detect or otherwise determine possible areas on the user's skin that may have blemishes. In particular, the segmentation model may generate a blemish map 202azmap that maps one or more possible blemishes or segments on the user's skin. A segment may comprise a grouping of one or more pixels in an image (e.g., image 202az) that have the same and / or similar pixel values ​​(e.g., the same and / or similar LAB and / or RGB values). For example, as shown in Figure 4C, a segmentation model is performed to determine a blemish map 202azmap, which includes various segments in the image that may have skin blemishes. This includes a blemish segment 202azmap1 which includes a pixel 202ap3 (having discoloration or otherwise a skin blemish), as described herein. Thus, in step 404, a computing instruction (e.g., of the imaging app) is executed by one or more processors (e.g., processor 104 or the processor of computing device 111c1), which in step 402 may cause one or more processors to generate a user-specific segmentation mapping of one or more blemishes in a portion of the user's skin area that is identifiable in the user's image.

[0076] An ensemble-based AI model's prediction or classification model may include any one or more of the following: a machine learning model, a regression equation, or a deep learning model. For example, a machine learning model may include a principal component model trained on multiple blemish maps (such as those output by a segmentation model) to determine the principal component (e.g., a feature such as a pixel feature) that has the highest predictive and / or discriminative value for accurately identifying a blemish or related blemish segment (e.g., 202azmap1) within a blemish map (e.g., 202azmap). Similarly, a machine learning model may include a regression equation with one or more independent variables having values ​​trained to identify the variable with the highest predictive and / or classification value for accurately identifying a blemish or related blemish segment (e.g., 202azmap1) within a blemish map (e.g., 202azmap). Furthermore, a neural network may be trained on a blemish map to constitute one or more nodes or hidden layers to identify the variable with the highest predictive and / or classification value for accurately identifying a blemish or related blemish segment (e.g., 202azmap1) within a blemish map (e.g., 202azmap).

[0077] In another embodiment, the RGB image shown in Figure 4C is converted into hemoglobin and melanin images (Figures 4D and 4E, respectively) using a regression model, a machine learning model (such as independent component analysis or principal component analysis), or a deep learning model (such as a generative neural network). Hemoglobin and melanin levels are measured by overlaying a spot map onto the predicted hemoglobin and melanin images. Spots are classified as either melanin spots or hemoglobin spots based on the ratio of hemoglobin to melanin. This can be further optimized by normalizing the melanin and hemoglobin values ​​of the spots against the melanin and hemoglobin values ​​of the surrounding skin.

[0078] When the instruction is executed by the processor, a prediction or classification model may further output a predicted or classified value indicating the blemish type (e.g., hemoglobin and / or melanin blemish type). For example, step 406 shows the classification of a hemoglobin blemish type, and Figure 4D illustrates image 202azhem having blemish 202azhem1 identified as a hemoglobin blemish type. In contrast, step 408 shows the classification of a melanin blemish type, and Figure 4E illustrates image 202azmel having blemish 202azmel1 identified as a melanin blemish type. In some embodiments, blemish type or blemish classification by other means may be identified by a blemish identifier (ID), where one set of blemish IDs (e.g., IDs 1-7) is a pixel value (e.g., RGB and / or L * a * b * Based on the pixel values ​​(e.g., RGB and / or L), a group of spots is selected to identify a variety of levels of skin pigmentation or hemoglobin spots of varying intensity, as identifiable within the pixel data, and another set of spot IDs (e.g., IDs 10-17) is selected to identify a group of spots that are identifiable within the pixel data as having different levels of pigmentation or intensity of hemoglobin spots, and another set of spot IDs (e.g., IDs 10-17) is selected to identify a group of spots that are identifiable within the pixel data as having different levels of pigmentation or intensity of hemoglobin spots, and a separate hemoglobin spots of varying intensity, based on the pixel values ​​(e.g., RGB and / or L * a * b * Based on the value, it is selected to identify a different group of spots that identify various levels of skin pigmentation or melanin spots of varying intensity as identifiable within the pixel data. Please understand that additional and / or different spot IDs can be used to detect or classify additional and / or different types of spots or discoloration on the user's skin.

[0079] Furthermore, with respect to Figures 4A to 4E, computing instructions may be further executed to determine at least one blemish classification (e.g., hemoglobin-type blemish or melanin-type blemish) selected from one or more blemish classifications of a skin-based learning model based on a predicted or classified value (e.g., blemish ID). Blemish types may be classified as inflammatory (red) blemishes indicating the presence of hemoglobin in the skin, and / or pigmented (brown) blemishes indicating the presence of melanin in the skin. In some embodiments, blemish types may be based on the detected ratio of hemoglobin to melanin presence. For example, in some embodiments, the predicted or classified value output by the skin-based learning model may be used to generate or update a chromophore image, where embodiments of pixel values ​​(e.g., indicating hemoglobin and / or melanin) are superimposed on each other, and the detected hemoglobin to melanin ratio may be determined from the pixels resulting in the chromophore image. In such embodiments, the blemish classification or type may be determined based on the dominance (e.g., percentage) of one type over the other. Additionally or alternatively, the classification may include hybrid types identified as having two or more types (e.g., both hemoglobin and melanin) present at the site of the blemish. In such embodiments, at least one blemish feature identifiable within the pixel data and / or one or more blemish classifications may be based on the skin's biological chromophore, containing one or more of eumelanin, pheomelanin, oxyhemoglobin, deoxyhemoglobin, bilirubin, or oxidized sebum.

[0080] One or more products may be recommended to treat the identified type of blemish; for example, an anti-inflammatory product may be recommended for hemoglobin-type blemishes. For melanin-type blemishes, a product for treating hyperpigmentation may be recommended.

[0081] Figure 5 illustrates exemplary user interfaces 502 rendered on the display screen 500 of a user computing device (e.g., user computing device 111c1) according to various embodiments disclosed herein. For example, as shown in the example of Figure 5, the user interface 502 may be implemented or rendered through an application (app) running on the user computing device 111c1. For example, as shown in the example of Figure 5, the user interface 502 may be implemented or rendered through a native app running on the user computing device 111c1. In the example of Figure 5, the user computing device 111c1 is the user computer device described for Figure 1, for example, 111c1 is illustrated as an Apple iPhone having a display screen 500 and implementing the Apple iOS operating system. The user computing device 111c1 may run one or more native applications (apps) on its operating system, including, for example, the imaging app described herein. Such native applications may be implemented or coded by the processor of the user computing device 111c1 in a computing language (e.g., SWIFT) that can be executed by the user computing device operating system (e.g., APPLE iOS) (e.g., as computing instructions).

[0082] Additionally, or alternatively, the user interface 502 may be implemented or rendered via a web interface, for example, via a web browser application such as the Safari and / or Google Chrome app, or other such web browsers.

[0083] As shown in the example in Figure 5, the user interface 502 includes a graphical representation of the user's skin (e.g., image 202az or a portion thereof). Image 202az may include an image of the user (or a graphical representation thereof) that includes pixel data (e.g., pixel data 202ap) of at least a portion of the skin area of ​​the user's skin, as described herein. In the example in Figure 5, the graphical representation of the user (e.g., image 202az) is annotated with one or more graphics (e.g., the area of ​​pixel data 202ap) or text renderings (e.g., text 202at) corresponding to various features identifiable within the pixel data that includes a portion of the user's skin area. For example, the area of ​​pixel data 202ap can be annotated or overlaid on the user's image (e.g., image 202az) to highlight areas or features identified within the pixel data (e.g., feature data and / or raw pixel data) by a skin-based learning model (e.g., skin-based learning model 108). In the example of Figure 5, the region of pixel data 202ap represents features indicating melanin spots (e.g., pixels near or at the location of 202ap3), as defined in pixel data 202ap, and may also represent other features shown in the region of pixel data 202ap, as described herein. In various embodiments, pixels identified as specific features (e.g., any of pixels 202ap1-202ap3) may be highlighted or otherwise annotated when rendered on the display screen 500.

[0084] A text rendering (e.g., text 202at) indicates a user-specific attribute or feature (e.g., the value "14" for pixel 202ap3), which may indicate that a pixel near or at the location of pixel 202ap3 has a blemish ID of 14 for skin coloring in that area. An ID value of 14 (e.g., on a scale of 11-17) indicates that the user has a mild or more severe color abnormality in a particular skin area (e.g., the skin area 202az) compared to the rest of the user's skin, and that the user is likely to benefit from using a product that improves skin texture and appearance (e.g., a product that normalizes blemishes and other skin discoloration). It should be understood that other text rendering types or values ​​are also envisioned in this specification. These text rendering types or values ​​may be rendered as blemish IDs such as melanin, hemoglobin, etc. Additionally, or alternatively, color values ​​can be used to indicate the degree or quality of a particular blemish ID by overlaying them on a graphical representation (e.g., image 202az) displayed on the user interface 502. For example, a high ID value of 17 or a low ID value of 11 (for example, a low RGB value and / or L * a * b * These include pixel values, etc. These IDs can be provided as raw values, absolute scores, percentage-based values, or as IDs. Additionally or alternatively, these IDs can be displayed with text or graphical indicators indicating whether the ID represents a positive result (e.g., low discoloration indicates low sun exposure or skin irritation), a negative result (e.g., high discoloration indicates excessive sun exposure or skin irritation), or an acceptable result (mean or acceptable value).

[0085] The user interface 502 may also include or render a user-specific blemish classification 510. In the embodiment of Figure 5, the user-specific blemish classification 510 includes a message 510m designed to show the user-specific blemish classification to the user, and a brief explanation of the reason for arriving at the user-specific blemish classification. As shown in the example of Figure 5, the message 512m indicates to the user that the user-specific blemish classification is "mild" (e.g., level 14), and further indicates that the user-specific blemish classification is due to excess melanin in a specified area of ​​the user's skin.

[0086] The user interface 502 may also include or render user-specific skin recommendations 512. For example, an imaging app may render at least one user-specific skin recommendation on the display screen of a computing device based on a user-specific blemish classification. In various embodiments, user-specific skin recommendations may include text recommendations, imaging-based recommendations, and / or virtual renderings of at least a portion of the user's skin area. For example, in the embodiment of Figure 5, the user-specific skin recommendation 512 includes a message 512m to the user, designed to address at least one identifiable feature in pixel data that includes a portion of the user's skin area. As shown in the example of Figure 5, the message 512m recommends to the user that they use a night face cream to help reduce dark spots. The night face cream product may be a low pH hydroxycinnamic acid (HCA) and niacinamide composition as described herein. Product recommendations can be made based on a blemish ID (e.g., value 14) that suggests the user's image depicts mild discoloration, and the night cream product is designed to address the discoloration detected or classified within the pixel data of image 202az, or the assumed discoloration based on the blemish ID or classification output by model 108. Product recommendations can be correlated with identified features within the pixel data, and the user computing device 111c1 and / or server 102 can be instructed to output a product recommendation when a feature (e.g., hypermelanin) is identified or classified.

[0087] The user interface 502 may also include or render a section for specific product recommendations 522 for a manufactured product 524r (e.g., a night face cream as described above). The product recommendations 522 may correspond to user-specific skin recommendations 512, as described above. For example, in the example of Figure 5, user-specific skin recommendations 512 may be displayed on the display screen 500 of the user's computing device 111c1 along with instructions (e.g., message 512m) to treat at least one identifiable feature (e.g., a mild spot ID 14 associated with melanin in a pixel near or at the location of 202ap3) contained in the pixel data (e.g., pixel data 202ap) of at least a portion of the pixel data of a skin area of ​​the user's skin (manufactured product 524r (e.g., a night face cream)) using the manufactured product (manufactured product 524r (e.g., a night face cream)).

[0088] As shown in Figure 5, the user interface 502 recommends a product (e.g., manufactured product 524r (e.g., night face cream)) based on user-specific skin recommendations 512. In the example in Figure 5, the output or analysis results of a skin image (e.g., image 202az) from a skin-based learning model (e.g., skin-based learning model 108), such as user-specific blemish classification 510 and / or its associated value (e.g., value 3) or associated pixel data (e.g., 202ap1, 202ap2, and / or 202ap3), and / or user-specific skin recommendations 512 can be used to generate or identify recommendations for the corresponding product. Such recommendations may include night face cream, exfoliants, moisturizers, moisturizing treatments, information to avoid excessive sun exposure, etc., and are intended to address user-specific issues detected in the pixel data by the skin-based learning model (e.g., skin-based learning model 108).

[0089] In the example in Figure 5, the user interface 502 displays or provides a skin-based learning model (e.g., skin-based learning model 108) and recommended products (e.g., manufactured products 524r) ​​determined by relevant image analysis of image 202az and its pixel data and various features. In the example in Figure 5, this is shown on the user interface 502 and annotated (524p).

[0090] The user interface 502 may further include selectable UI buttons 524s that enable a user (e.g., the user in image 202az) to select a corresponding product (e.g., a manufactured product 524r) ​​for purchase or shipment. In some embodiments, the selection of a selectable UI button 524s may result in a recommended product being shipped to the user (e.g., user 202a in the image) and / or a third party being notified that the individual is interested in that product. For example, the user computing device 111c1 and / or the imaging server 102 may initiate the shipment of a manufactured product 524r (e.g., a night face cream) to the user based on the user-specific blemish classification 510 and / or the user-specific skin recommendation 512. In such embodiments, the product may be packaged and shipped to the user.

[0091] In various embodiments, a graphical display (e.g., image 202az), graphical annotations (e.g., pixel data area 202ap), text annotations (e.g., text 202at), user-specific blemish classification 510, and user-specific skin recommendations 512 may be transmitted to the user computing device 111c1 via a computer network (e.g., from the pixel server 102 and / or one or more processors) and rendered on the display screen 500. In other embodiments, user-specific images are not transmitted to the pixel server, and the user-specific blemish classification 510 and user-specific skin recommendations 512 (and / or product-specific recommendations) are generated locally by a skin-based learning model (e.g., skin-based learning model 108) running and / or implemented on the user's mobile device (e.g., user computing device 111c1) and rendered on the mobile device's display screen 500 by the mobile device's processor.

[0092] In some embodiments, one or more of the following may be rendered in real time or near real time (e.g., rendered locally on a display screen 500) while or after receiving an image containing the skin area of ​​the user's skin: graphical display (e.g., image 202az), graphical annotation (e.g., pixel data area 202ap), text annotation (e.g., text 202at), user-specific blemish classification 510, user-specific skin recommendations 512, and / or product recommendations 522. In embodiments where the image is analyzed by the imaging server 102, the image may be transmitted and analyzed by the imaging server 102 in real time or near real time.

[0093] In some embodiments, a user can provide a new image, which can be sent to the pixel server 102 for updating, retraining, or reanalysis by the skin-based learning model 108. In other embodiments, the new image can be received locally on the computing device 111c1 and analyzed on the computing device 111c1 by the skin-based learning model 108.

[0094] In addition, as shown in the example in Figure 5, the user may select a selectable button 512i to reanalyze a new image (e.g., locally on the computing device 111c1 or remotely on the imaging server 102). The selectable button 512i allows the user interface 502 to prompt the user to attach the new image for analysis. The imaging server 102 and / or user computing devices such as the user computing device 111c1 may receive a new image containing pixel data of at least some of the skin regions of the user's skin. The new image may be captured by the imaging device. The new image (e.g., similar to image 202az) may contain pixel data of at least some of the skin regions of the user's skin. A skin-based learning model (e.g., skin-based learning model 108) running in the memory of the computing device (e.g., imaging server 102) can analyze the new image captured by the imaging device and determine the image classification of the user's skin regions. A computing device (e.g., imaging server 102) can generate a new user-specific blemish classification and / or a new user-specific skin recommendation based on a comparison between an image and a second image, or between a classification of the user's skin area and a second classification, relating to at least one feature identifiable within the pixel data of the new image. For example, the new user-specific blemish classification may include a new graphic display including graphics and / or text (e.g., displaying a new skin blemish ID value (e.g., 11) after the user has used night face cream). The new user-specific blemish classification may include additional blemish classifications, such as the user successfully reducing melanin detected in the pixel data of the new image by using night face cream. Comments may include additional features detected by the user in the pixel data, such as additional blemishes that need to be corrected by applying an additional product (e.g., moisturizing oil).

[0095] In various embodiments, new user-specific blemish classifications and / or new user-specific skin recommendations may be transmitted from server 102 to the user's user computing device (e.g., user computing device 111c1) via a computer network for rendering on the user's user computing device's display screen 500.

[0096] In other embodiments, the user's new image is not sent to the pixel server, and the new user-specific blemish classification and / or new user-specific skin recommendations (and / or product-specific recommendations) are generated locally by a skin-based learning model (e.g., skin-based learning model 108a) running and / or implemented on the user's mobile device (e.g., user computing device 111c1), and can be rendered on the mobile device's display screen (e.g., user computing device 111c1) by the mobile device's processor.

[0097] The nature of this disclosure The following embodiments are provided as examples of the disclosure herein and are not intended to limit the scope of this disclosure. 1. A digital imaging and artificial intelligence-based system configured to analyze pixel data of a user's skin image and generate one or more user-specific skin spot classifications, comprising: one or more processors; a skin application (app) including computing instructions configured to run on one or more processors; and a skin-based learning model accessible by the imaging app and trained using pixel data of multiple training images depicting each individual's skin, wherein the skin-based learning model is configured to output one or more spot classifications corresponding to one or more spot features in each individual's skin region. A digital imaging and artificial intelligence-based system in which, when computing instructions for an imaging application are executed by one or more processors, causes one or more processors to receive a user image which includes a digital image captured by an imaging device and pixel data of at least a portion of the user's skin area, a skin-based learning model analyzes the image captured by the imaging device and determines at least one blemish classification of the user's skin, selected from one or more blemish classifications of the skin-based learning model, and generates user-specific skin recommendations designed to address at least one identifiable blemish feature in the pixel data which includes a portion of the user's skin area, based on the at least one blemish classification of the user's skin. 2. A digital imaging and artificial intelligence-based system according to Embodiment 1, wherein at least one blemish feature or one or more blemish classifications identifiable within the pixel data are based on the biological chromophore of the skin, comprising one or more of eumelanin, pheomelanin, oxyhemoglobin, deoxyhemoglobin, bilirubin, or oxidized sebum. 3. A digital imaging and artificial intelligence-based system according to one or more embodiments 1 to 2, wherein one or more spot classifications include one or more of (1) hemoglobin type classifications or (2) melanin type classifications. 4. A digital imaging and artificial intelligence-based system according to any one or more of embodiments 1 to 3, wherein an image calibration algorithm is applied to each of multiple training images to modify the images to enhance spot classification, and when computing instructions for the imaging application are executed by one or more processors, one or more processors are further instructed to apply the image calibration algorithm to the user's image before analyzing the user's image using a skin-based learning model. 5. A digital imaging and artificial intelligence-based system according to any one or more embodiments 1 to 3, wherein the skin-based learning model is an ensemble-based AI model comprising (i) a segmentation model configured to generate segmentation mappings of one or more blemishes in skin regions of an image, and (ii) a prediction or classification model configured to analyze the pixel data of the segmentation mappings of one or more blemishes, wherein when a computing instruction of an imaging application is executed by one or more processors, one or more processors are further caused to generate user-specific segmentation mappings of one or more blemishes in a portion of the user's skin region that is identifiable in the user's image, the prediction or classification model outputs a prediction or classification value indicating the blemish type, and based on the prediction or classification value, at least one blemish classification selected from one or more blemish classifications of the skin-based learning model. 6. A digital imaging and artificial intelligence-based system according to any one or more embodiments 1 to 4, wherein each of the multiple training images or the user's image includes at least one cropped image depicting a skin region having a single instance of a blemish feature. 7. A digital imaging and artificial intelligence-based system according to any one of embodiments 1 to 5, wherein one or more images from a plurality of training images or user images include multiple angles or viewpoints depicting the skin area of ​​the respective individual or user. 8. A digital imaging and artificial intelligence-based system according to any one or more embodiments 1 to 6, wherein the computing instructions of the imaging application are executed by one or more processors, and one or more processors further cause the computing device to render at least one user-specific skin recommendation based on user-specific blemish classification on the display screen of the computing device. 9. A digital imaging and artificial intelligence-based system according to any one or more embodiments 1 to 7, wherein at least one user-specific skin recommendation is displayed on the display screen of a computing device, along with instructions for treating at least one identifiable spot feature in pixel data that includes a portion of the user's skin area. 10. A digital imaging and artificial intelligence-based system according to any one or more embodiments 1 to 8, wherein at least one user-specific skin recommendation includes text recommendations, image-based recommendations, or virtual renderings of at least a portion of the user's skin area. 11. A digital imaging and artificial intelligence-based system according to any one or more embodiments 1 to 9, wherein at least one user-specific skin recommendation is rendered on a display screen in real time or near real time while or after receiving the user's image. 12. A digital imaging and artificial intelligence-based system according to Embodiment 7, wherein at least one user-specific stain recommendation includes a product recommendation relating to a manufactured product. 13. The digital imaging and artificial intelligence-based system according to embodiment 11, wherein at least one user-specific skin recommendation is displayed on the display screen of a computing device, along with instructions to treat at least one identifiable spot feature in pixel data including at least a portion of the user's skin area with a manufactured product. 14. The digital imaging and artificial intelligence-based system according to aspect 11, wherein these computing instructions further cause one or more processors to initiate shipping a manufactured product to a user based on at least one user-specific skin recommendation. 15. A digital imaging and artificial intelligence-based system according to aspect 11, wherein a computing instruction causes one or more processors to generate an image-modified image that depicts what the user's skin area is expected to look like after at least one blemish feature has been treated with a manufactured product, and to render the modified image on the display screen of the computing device. 16. A digital imaging and artificial intelligence-based system according to any one or more embodiments 1 to 14, wherein the computing instructions of the imaging application are executed by one or more processors, and further cause one or more processors to generate a skin type code determined based on a user-specific blemish classification, which is designed to address at least one blemish feature identifiable within pixel data, including a portion of the user's skin area. 17. A digital imaging and artificial intelligence-based system according to any one or more embodiments 1 to 15, wherein a computing instruction causes one or more processors, and further, one or more memories communicably coupled to one or more processors, to record an image of the user captured by an imaging device at a first time in order to track changes in the user's skin area over time; to receive a second image of the user captured by an imaging device, which includes pixel data of at least a portion of the user's skin area, at a second time; to analyze the second image captured by the imaging device using a skin-based learning model at a second time; to determine a second image classification of the user's skin area selected from one or more image classifications of the skin-based learning model; and to generate a new user-specific blemish classification relating to at least one blemish feature identifiable within the pixel data of the second image, which includes at least a portion of the user's skin area, or its absence, based on a comparison of the image of the user's skin area with the second image and / or the image classification and the second image classification. 18. A digital imaging and artificial intelligence-based system according to any one of embodiments 1 to 16, wherein the skin-based learning model is an AI-based model trained using at least one artificial intelligence (AI) algorithm. 19. A digital imaging and artificial intelligence-based system according to any one or more embodiments 1 to 17, wherein one or more blemish features in the skin regions of multiple training images differ based on one or more demographic attributes or ethnicity of each individual user, and the user-specific blemish classification of a user is generated by a skin-based learning model based on the user's ethnicity or demographic attribute values. 20. A digital imaging and artificial intelligence-based system according to any one or more embodiments 1 to 19, wherein the skin-based learning model is further trained using the user's demographic and environmental data, and at least one blemish classification generated by the skin-based learning model is further based on the user's demographic and environmental data provided by the user. 21. A digital imaging and artificial intelligence-based system according to any one or more embodiments 1 to 20, wherein at least one of the one or more processors includes a processor for a mobile device, and the imaging device includes a digital camera for a mobile device. 22. A digital imaging and artificial intelligence-based system according to any one or more embodiments 1 to 21, comprising: a server processor of a server, the server being communicably coupled to a computing device via a computer network, and an imaging application comprising: a server application portion configured to run on one or more processors of the server, and a computing device application portion configured to run on one or more processors of the computing device, wherein the server application portion is configured to communicate with the computing device application portion, and the server application portion is configured to implement one or more of the following: (1) receiving images captured by the imaging device, (2) determining at least one spot classification of a user's skin area, (3) generating user-specific spot classifications, and / or (4) sending user-specific recommendations to the computing device application portion. 23. A digital imaging and artificial intelligence-based method for analyzing pixel data of a user's skin image to generate one or more user-specific skin spot classifications, the digital imaging and artificial intelligence-based method comprising: receiving a user image, which is a digital image captured by an imaging device and includes pixel data of at least a portion of the user's skin region, in one or more processors; analyzing the image captured by the imaging device with a skin-based learning model, which is run on one or more processors, trained using pixel data of multiple training images depicting each individual's skin, and configured to output one or more spot classifications corresponding to one or more spot features in each individual's skin region, and determining at least one spot classification of the user's skin to be selected from one or more spot classifications of the skin-based learning model; and generating a user-specific skin recommendation, designed to address at least one identifiable spot feature in the pixel data including a portion of the user's skin region, based on at least one spot classification of the user's skin, in one or more processors. 24. The digital imaging and artificial intelligence-based method according to aspect 23, wherein at least one blemish feature or one or more blemish classifications identifiable within the pixel data are based on the biological chromophore of the skin, comprising one or more of eumelanin, pheomelanin, oxyhemoglobin, deoxyhemoglobin, bilirubin, or oxidized sebum. 25. A digital imaging and artificial intelligence-based method according to any one or more embodiments 23 to 24, wherein one or more spot classifications include one or more of (1) hemoglobin type classifications or (2) melanin type classifications. 26. A digital imaging and artificial intelligence-based method according to one or more of embodiments 23 to 25, wherein an image calibration algorithm is applied to each of a plurality of training images to modify the images to enhance spot classification, and when computing instructions of the imaging application are executed by one or more processors, one or more processors are further instructed to apply the image calibration algorithm to the user's images before analyzing the user's images using a skin-based learning model. 27. A digital imaging and artificial intelligence-based method according to one or more embodiments of 23 to 26, wherein the skin-based learning model is an ensemble-based AI model comprising (i) a segmentation model configured to generate segmentation mappings of one or more blemishes in skin regions of an image, and (ii) a prediction or classification model configured to analyze pixel data of the segmentation mappings of one or more blemishes, wherein when a computing instruction of an imaging application is executed by one or more processors, one or more processors are further caused to generate user-specific segmentation mappings of one or more blemishes in a portion of the user's skin region identifiable in the user's image, the prediction or classification model outputs a prediction or classification value indicating the blemish type, and based on the prediction or classification value, at least one blemish classification selected from one or more blemish classifications of the skin-based learning model. 28. A digital imaging and artificial intelligence-based method according to any one or more of embodiments 23 to 27, wherein each of the multiple training images or the user's image includes at least one cropped image depicting a skin region having a single instance of a blemish feature. 29. A digital imaging and artificial intelligence-based method according to any one of embodiments 23 to 28, wherein one or more images from a plurality of training images or user images include multiple angles or viewpoints depicting the respective individual or user's skin area. 30. A digital imaging and artificial intelligence-based method according to any one of embodiments 23 to 29, wherein when a computing instruction of an imaging application is executed by one or more processors, one or more processors further cause one or more processors to render at least one user-specific skin recommendation based on user-specific blemish classification on the display screen of the computing device. 31. The digital imaging and artificial intelligence-based method according to embodiment 30, wherein at least one user-specific skin recommendation is displayed on the display screen of a computing device along with instructions for treating at least one identifiable spot feature in pixel data including a portion of the user's skin area. 32. The digital imaging and artificial intelligence-based method according to embodiment 30, wherein at least one user-specific skin recommendation includes a text recommendation, an image-based recommendation, or a virtual rendering of at least a portion of the user's skin area. 33. The digital imaging and artificial intelligence-based method according to aspect 30, wherein at least one user-specific skin recommendation is rendered on a display screen in real time or near real time while or after receiving the user's image. 34. The digital imaging and artificial intelligence-based method according to embodiment 30, wherein at least one user-specific stain recommendation includes a product recommendation for a manufactured product. 35. The digital imaging and artificial intelligence-based method according to aspect 34, wherein at least one user-specific skin recommendation is displayed on the display screen of a computing device, along with instructions to treat at least one identifiable spot feature in pixel data including a portion of the user's skin area with a manufactured product. 36. The digital imaging and artificial intelligence-based method according to aspect 34, wherein these computing instructions further cause one or more processors to initiate shipping a manufactured product to a user based on at least one user-specific skin recommendation. 37. The digital imaging and artificial intelligence-based method according to aspect 34, wherein a computing instruction causes one or more processors to generate an image-modified image that depicts what the user's skin area is expected to look like after at least one blemish feature has been treated with a manufactured product, and to render the modified image on the display screen of the computing device. 38. A digital imaging and artificial intelligence-based method according to any one of embodiments 23 to 37, wherein, when the computing instructions of the imaging application are executed by one or more processors, one or more processors further cause one or more processors to generate a skin type code determined based on a user-specific blemish classification, which is designed to address at least one blemish feature identifiable within pixel data including at least a portion of the user's skin area. 39. A digital imaging and artificial intelligence-based method according to any one or more embodiments 23 to 38, wherein a computing instruction further causes one or more processors to record an image of the user captured by an imaging device in one or more memories communicably coupled to one or more processors in order to track changes in the user's skin area over time; to receive a second image of the user captured by an imaging device in a second time, which includes pixel data of at least a portion of the user's skin area; to analyze the second image captured by the imaging device using a skin-based learning model in a second time, to determine a second image classification of the user's skin area selected from one or more image classifications of the skin-based learning model; and to generate a new user-specific blemish classification relating to at least one blemish feature identifiable in the pixel data of the second image including at least a portion of the user's skin area, or its absence, based on a comparison of the image of the user's skin area with the second image and / or the image classification and the second image classification. 40. A digital imaging and artificial intelligence-based method according to any one of embodiments 23 to 39, wherein the skin-based learning model is an AI-based model trained using at least one artificial intelligence (AI) algorithm. 41. The digital imaging and artificial intelligence-based method according to aspect 40, wherein one or more blemish features of skin regions in multiple training images differ based on one or more demographic attributes or ethnicity of each individual user, and the user-specific blemish classification of a user is generated by a skin-based learning model based on the user's ethnicity or demographic attribute values. 42. A digital imaging and artificial intelligence-based method according to any one or more embodiments 23 to 21, wherein the skin-based learning model is further trained using the user's demographic and environmental data, and at least one spot classification generated by the skin-based learning model is further based on the user's demographic and environmental data provided by the user. 43. A digital imaging and artificial intelligence-based method according to any one or more embodiments 23 to 41, wherein at least one of the one or more processors includes a processor for a mobile device, and the imaging device includes a digital camera for a mobile device. 44. A digital imaging and artificial intelligence-based method according to any one or more embodiments 23 to 43, comprising: 1 or more processors comprising a server processor of a server, the server being communicably coupled to a computing device via a computer network, and an imaging application comprising: a server application portion configured to run on one or more processors of the server, and a computing device application portion configured to run on one or more processors of the computing device, wherein the server application portion is configured to communicate with the computing device application portion, and the server application portion is configured to implement one or more of the following: (1) receiving images captured by the imaging device, (2) determining at least one spot classification of a user's skin area, (3) generating user-specific spot classifications, and / or (4) sending user-specific recommendations to the computing device application portion. 45. A tangible, non-temporary, computer-readable medium that stores instructions for analyzing pixel data of a user's skin image in order to generate one or more user-specific skin spot classifications, wherein, when an instruction is executed by one or more processors, the instructions cause one or more processors to receive a user image, which includes a digital image captured by an imaging device, the image including pixel data of at least a portion of the user's skin region, and to analyze the image captured by the imaging device using a skin-based learning model that is trained using pixel data of multiple training images depicting each individual's skin and configured to output one or more spot classifications corresponding to one or more spot features in each individual's skin region, to determine at least one spot classification of the user's skin, at least one spot classification selected from one or more spot classifications of the skin-based learning model, and to generate a user-specific skin recommendation designed to address at least one identifiable feature in the pixel data including a portion of the user's skin region, based on at least one spot classification of the user's skin.

[0098] Additional considerations While the disclosure herein provides a detailed description of numerous different embodiments, it should be understood that the legal scope of this specification is defined by the claims wording set forth at the end of this Patent and its Equivalents. This detailed description should be interpreted as illustrative only and does not describe all possible embodiments, as it would be impractical to do so. Numerous alternative embodiments may be implemented using either the current art or art developed after the filing date of this Patent, but such embodiments still fall within the scope of the claims.

[0099] The following additional considerations apply to the preceding discussion. Throughout this specification, multiple examples may implement a component, operation, or structure described as a single example. While individual operations of one or more methods are illustrated and described as separate operations, one or more of these operations may be performed simultaneously, and they do not need to be performed in the order in which they are illustrated. Structures and functions presented as separate components in exemplary configurations may be implemented as combined structures or components. Similarly, structures and functions presented as single components may also be implemented as separate components. These and other variations, modifications, additions, and improvements fall within the scope of this specification.

[0100] In addition, certain embodiments described herein include logic or several routines, subroutines, applications, or instructions. These may constitute either software (e.g., code embodied on a machine-readable medium or in a transmitted signal) or hardware. In hardware, routines, etc., are tangible units capable of performing specific operations and may be configured or arranged in a particular manner. In exemplary embodiments, one or more computer systems (e.g., standalone, client, or server computer systems) or one or more hardware modules of a computer system (e.g., a processor or a group of processors) may be configured by software (e.g., an application or application portion) as hardware modules that operate to perform specific operations as described herein.

[0101] Various operations of the exemplary methods described herein may be performed, at least in part, by one or more processors that are temporarily (e.g., by software) or permanently configured to perform the relevant operations. Whether temporarily or permanently configured, such processors may constitute a processor implementation module that operates to perform one or more operations or functions. The modules referred to herein may include processor implementation modules in some exemplary embodiments.

[0102] Similarly, any methods or routines described herein can be at least partially processor-implemented. For example, at least some of the operations of a particular method may be performed by one or more processors or processor-implementation hardware modules. The specific performance of those operations may be distributed across one or more processors and may reside not only within a single machine but also across multiple machines. In some exemplary embodiments, one or more processors may be located in a single location, while in other embodiments, the processors may be distributed across several locations.

[0103] The specific performance characteristics of these operations can be distributed across one or more processors and can reside not only within a single machine but also deployed across multiple machines. In some exemplary embodiments, one or more processors or processor implementation modules may be located in a single geographical location (e.g., a home environment, an office environment, or a server farm). In other embodiments, one or more processors or processor implementation modules may be distributed across multiple geographical locations.

[0104] This detailed description should be interpreted as illustrative only, and does not describe all possible embodiments, as it would be impractical, if not impossible, to describe all possible embodiments. Those skilled in the art may implement numerous alternative embodiments using either the current art or art developed after the filing date of this application.

[0105] Those skilled in the art will recognize that a wide variety of modifications, changes, and combinations can be made with respect to the above embodiments without departing from the scope of the present invention, and that such modifications, changes, and combinations should be considered to fall within the scope of the concept of the present invention.

[0106] The claims at the end of this patent application are not intended to be construed under Section 112(f) of the United States Patent Act unless the conventional means plus functional language is explicitly enumerated, for example, unless the language “means for” or “process for” is explicitly enumerated in the claims. The systems and methods described herein are intended to improve computer functions and improve the functions of conventional computers.

[0107] The dimensions and values ​​disclosed herein should not be understood as being strictly limited to the exact numerical values ​​listed. Instead, unless otherwise specified, each such dimension is intended to mean both the listed value and the functionally equivalent range encompassing that value. For example, a dimension disclosed as "40 mm" is intended to mean "approximately 40 mm."

[0108] All documents referenced herein, including any patents or patent applications that are cross-referenced or related, and any patent applications or patents on which this application claims priority or benefit thereof, are incorporated herein by reference in their entirety, unless expressly excluded or otherwise limited. No reference to any document shall be deemed prior art to any invention disclosed or claimed herein, nor shall any such invention be taught, suggested, or disclosed, either alone or in combination with any one or more other references. Furthermore, to the extent that any meaning or definition of a term in this document conflicts with any meaning or definition of the same term in any document incorporated by reference, the meaning or definition given to that term in this document shall prevail.

[0109] While specific embodiments of the present invention have been illustrated and described, it will be apparent to those skilled in the art that various other changes and modifications can be made without departing from the spirit and scope of the invention. Therefore, it is intended that all such changes and modifications within the scope of the invention be covered in the appended claims.

Claims

1. A digital imaging and artificial intelligence-based system configured to analyze pixel data of a user's skin image and generate one or more user-specific skin spot classifications, wherein the digital imaging and artificial intelligence-based system is configured to analyze pixel data of a user's skin image and generate one or more user-specific skin spot classifications, One or more processors, An imaging application (app) including computing instructions configured to run on one or more processors, A skin-based learning model, accessible by the aforementioned imaging application and trained using pixel data from multiple training images depicting each individual's skin, is configured to output one or more blemish classifications corresponding to one or more blemish features in each individual's skin region. When the computing instructions of the imaging application are executed by one or more processors, the one or more processors: The system receives an image of the user, which includes a digital image captured by the imaging device and includes pixel data of at least a portion of the user's skin area. The skin-based learning model analyzes the image captured by the imaging device and determines at least one blemish classification of the user's skin, selected from the one or more blemish classifications of the skin-based learning model, and A digital imaging and artificial intelligence-based system that generates user-specific skin recommendations designed to address at least one identifiable blemish feature within pixel data including a portion of the user's skin region, based on the at least one blemish classification of the user's skin.

2. The digital imaging and artificial intelligence-based system according to claim 1, wherein the at least one blemish feature or one or more blemish classifications identifiable within the pixel data are based on a biological chromophore of the skin comprising one or more eumelanin, pheomelanin, oxyhemoglobin, deoxyhemoglobin, bilirubin, or oxidized sebum.

3. The digital imaging and artificial intelligence-based system according to claim 1 or 2, wherein the one or more spot classifications include one or more of (1) hemoglobin type classifications or (2) melanin type classifications.

4. An image calibration algorithm is applied to each of the aforementioned training images to modify the images in order to enhance spot classification, and When the computing instructions of the imaging application are executed by one or more processors, the one or more processors further: A digital imaging and artificial intelligence-based system according to any one of claims 1 to 3, wherein the image calibration algorithm is applied to the user's image before the user's image is analyzed using the skin-based learning model.

5. The skin-based learning model is an ensemble-based AI model comprising (i) a segmentation model configured to generate segmentation mappings of one or more blemishes in the skin region of an image, and (ii) a prediction or classification model configured to analyze the pixel data of the segmentation mappings of one or more blemishes, and When the computing instructions of the imaging application are executed by one or more processors, the one or more processors further: To generate a user-specific segmentation mapping of one or more blemishes within a portion of the user's skin area that is identifiable in the user's image, The aforementioned prediction or classification model outputs a predicted or classified value indicating the stain type, and A digital imaging and artificial intelligence-based system according to any one of claims 1 to 4, which determines the at least one spot classification selected from the one or more spot classifications of the skin-based learning model based on the predicted value or classification value.

6. A digital imaging and artificial intelligence-based system according to any one of claims 1 to 5, wherein each of the one or more images from the plurality of training images or the user's image includes at least one cropped image depicting the skin region having a single instance of a blemish feature.

7. A digital imaging and artificial intelligence-based system according to any one of claims 1 to 6, wherein each of the one or more images from the plurality of training images or the image of the user includes a plurality of angles or viewpoints depicting the skin area of ​​the respective individual or the user.

8. When the computing instructions of the imaging application are executed by one or more processors, the one or more processors further: A digital imaging and artificial intelligence-based system according to any one of claims 1 to 7, which renders at least one user-specific skin recommendation based on the user-specific blemish classification on the display screen of a computing device.

9. The digital imaging and artificial intelligence-based system according to claim 8, wherein the at least one user-specific skin recommendation is displayed on the display screen of the computing device, along with instructions for processing the at least one blemish feature identifiable in the pixel data including the portion of the user's skin area.

10. The digital imaging and artificial intelligence-based system according to claim 8 or 9, wherein the at least one user-specific skin recommendation includes a text recommendation, an image-based recommendation, or a virtual rendering of at least a portion of the user's skin area.

11. The digital imaging and artificial intelligence-based system according to any one of claims 8 to 10, wherein the at least one user-specific skin recommendation is rendered on the display screen in real time or near real time while the user's image is being received or after it has been received.

12. The digital imaging and artificial intelligence-based system according to any one of claims 8 to 11, wherein the at least one user-specific stain recommendation includes a product recommendation for a manufactured product.

13. The computing instruction further provides the one or more processors with: The image modified based on the aforementioned image is generated to depict what the user's skin area is expected to look like after the at least one blemish feature has been processed with the manufactured product. The digital imaging and artificial intelligence-based system according to claim 12, wherein the modified image is rendered on the display screen of the computing device.

14. The computing instruction further provides the one or more processors with: One or more memories, which are communicably coupled to one or more processors, record the image of the user captured by the imaging device at a first time in order to track changes in the user's skin area over time. At a second time, the imaging device receives a second image of the user, which includes pixel data of at least a portion of the user's skin area. During the second time, the second image captured by the imaging device is analyzed by the skin-based learning model, and the second image classification of the user's skin region selected from the one or more image classifications of the skin-based learning model is determined, and A digital imaging and artificial intelligence-based system according to claim 1, which generates a new user-specific blemish classification relating to at least one blemish feature or absence thereof that can be identified in the pixel data of the second image including at least a portion of the user's skin area, based on a comparison of the image of the user's skin area and the second image and / or the image classification and the second image classification.

15. The aforementioned skin-based learning model is an AI-based model trained using at least one artificial intelligence (AI) algorithm. The one or more spot features in the skin region of the plurality of training images differ based on one or more user demographic attributes or ethnicity of each individual, and The digital imaging and artificial intelligence-based system according to claim 1, wherein the user's unique blemish classification is determined by the skin-based learning model based on the user's ethnicity or demographic attribute values.

Images