KR-20260067774-A - DEVICE AND METHOD FOR ARTIFICIAL INTELLIGENCE SKINCARE SERVICE PLATFORM BASED ON SELFIE VIDEOS

KR20260067774AKR 20260067774 AKR20260067774 AKR 20260067774AKR-20260067774-A

Abstract

The present invention relates to an artificial intelligence skincare service platform device and method based on selfie images. The device comprises: a selfie image input unit that selects a face image for skin analysis through face recognition of a selfie image; an image preprocessing unit that generates a light-neutral preprocessed face image through preprocessing of the face image for skin analysis; a time-series skin analysis unit that analyzes short-term skin changes by extracting skin features through inputting the preprocessed face image into a first skin analysis network and analyzes long-term skin changes by sequentially inputting the skin features into a second skin analysis network in a time-series order; a contextual skin prediction unit that extracts environmental factors through the selfie image, analyzes contextual lifestyle patterns through tracking of the environmental factors, and predicts future skin changes through the correlation between the long-term skin changes and the contextual lifestyle patterns; and a reporting transformer unit that generates a report on the short-term skin changes, the long-term skin changes, and the prediction of future skin changes based on the contextual lifestyle patterns, and processes user queries and responses.

Inventors

정우주
신명진
박수아
강우진

Assignees

주식회사 인디제이

Dates

Publication Date: 20260513
Application Date: 20241106

Claims (8)

A selfie video input unit that selects face videos for skin analysis through face recognition of selfie videos; An image preprocessing unit that generates a light-neutral preprocessed face image through preprocessing of the face image for skin analysis; A time-series skin analysis unit that analyzes short-term skin changes by inputting the above-mentioned preprocessed face image into a first skin analysis network to extract skin features, and analyzes long-term skin changes by sequentially inputting the above-mentioned skin features into a second skin analysis network in a time-series order; A contextual skin prediction unit that extracts environmental factors through the above-mentioned selfie video, analyzes contextual lifestyle patterns through tracking the above-mentioned environmental factors, and predicts future skin changes through the correlation between the above-mentioned long-term skin changes and the above-mentioned contextual lifestyle patterns; and A selfie video-based artificial intelligence skincare service platform device comprising a reporting transformer unit that generates reports on the prediction of short-term skin changes, long-term skin changes, and future skin changes based on the above contextual lifestyle patterns, and processes user queries and responses.
In claim 1, the selfie video input unit A selfie video-based artificial intelligence skincare service platform device characterized by automatically identifying the selfie video in advance among the videos in the photo library of a user terminal.
In claim 1, the image preprocessing unit A selfie video-based artificial intelligence skincare service platform device characterized by generating a copy image of the face image for skin analysis and generating the preprocessed face image through format standardization, noise removal, and clarity improvement of the copy image.
In paragraph 1, the time series skin analysis unit The above first skin analysis network is composed of a vision transformer model that receives the above preprocessed face image as input and a face skin trouble analysis model that receives the above skin features as input, and A selfie video-based artificial intelligence skincare service platform device characterized by the facial skin trouble analysis model receiving previously learned acne-specialized, wrinkle-specialized, pigmentation-specialized, and pore-specialized analyses through transfer learning, and selecting at least one of the specialized models based on the skin characteristics to analyze user-customized short-term skin changes.
In paragraph 4, the time-series skin analysis unit The above second skin analysis network is configured as an LSTM (Long Short-Term Memory) network, and A selfie video-based artificial intelligence skincare service platform device characterized by analyzing long-term skin changes by extracting periodic skin change patterns according to the input order of the skin features through the above LSTM network.
In paragraph 1, the contextual skin prediction unit A selfie video-based artificial intelligence skincare service platform device characterized by extracting the shooting location and shooting time from the metadata of the selfie video to identify user activity and seasonal elements, analyzing the contextual lifestyle pattern, and reflecting the contextual lifestyle pattern in the long-term skin change to predict the future skin change.
In paragraph 1, the reporting transformer part Automatically generates the above report focusing on explanations of skin change trends and skin care based on major changes exceeding a specific standard in the above short-term skin changes, the above long-term skin changes, and the above predictions of future skin changes. A selfie video-based artificial intelligence skincare service platform device characterized by performing skincare consultation by combining a user question-and-answer module with the above report.
In a selfie video-based artificial intelligence skincare service platform method performed on a selfie video-based artificial intelligence skincare service platform device, A selfie video input step for selecting face videos for skin analysis through face recognition of selfie videos; An image preprocessing step for generating a light-neutral preprocessed face image through preprocessing of the face image for skin analysis; A time-series skin analysis step in which the above-mentioned preprocessed face image is input into a first skin analysis network to extract skin features and analyze short-term skin changes, and the above-mentioned skin features are sequentially input into a second skin analysis network in a time-series order to analyze long-term skin changes; A contextual skin prediction step that extracts environmental factors through the above-mentioned selfie video, analyzes contextual lifestyle patterns through tracking the above-mentioned environmental factors, and predicts future skin changes through the correlation between the above-mentioned long-term skin changes and the above-mentioned contextual lifestyle patterns; and A method for a selfie video-based artificial intelligence skincare service platform comprising a reporting transformer step that generates reports on the prediction of short-term skin changes, long-term skin changes, and future skin changes based on the above contextual lifestyle patterns, and processes user queries and responses.

Description

Device and Method for Artificial Intelligence Skincare Service Platform Based on Selfie Videos The present invention relates to artificial intelligence skincare service technology, and more specifically, to a selfie video-based artificial intelligence skincare service platform device and method capable of selecting a face image suitable for skin analysis from a selfie video, and tracking skin changes through deep learning-based analysis after lighting-neutral preprocessing. In the field of skincare, various AI-based solutions are emerging as effective tools for diagnosing skin conditions and suggesting care plans. These AI-based solutions primarily operate by analyzing photos taken at specific points in time to evaluate skin condition and recommend general care methods. However, this approach has several significant limitations and requires improvement. Existing AI-based skincare systems adopt a method of analyzing skin condition using photos taken at a specific point in time, which leads them to rely solely on images intentionally taken by the user. Such one-time analysis makes it difficult to track continuous and long-term changes in skin condition. Furthermore, because they use only limited data, there is a problem in that they fail to adequately reflect the long-term impact of environmental factors or lifestyle patterns on the skin. Analysis methods utilizing only images from a specific point in time reveal limitations in comprehensively understanding overall changes in the skin. Furthermore, current AI-based skincare technologies often fail to adequately reflect an individual user's skin condition and lifestyle, frequently providing generalized advice. By proposing standardized management plans without considering individual user characteristics, personalized care is lacking, making it difficult to achieve effective skin care. Additionally, due to fragmented services where skin analysis, product recommendations, and specialist consultations are provided separately, users find it difficult to obtain a consistent and integrated skincare experience, leading to a complex management process. Korean Patent Publication No. 10-2020-0063299 (June 5, 2020) provides a method for learning patterns inherent in skin condition, products used, and environmental information through unsupervised learning and recommending products based on big data. A skincare recommendation service utilizing big data-based AI machine learning sends a Baumann skin type questionnaire to the user. Upon receiving the results, it diagnoses the user's skin type and collects product data—such as basic skincare, sun care, and cleansing products—along with environmental data including gender, age, occupation, and diet. This data is then parsed and classified. Subsequently, based on the classified identifiers, it extracts reference data with high similarity from the big data and recommends products containing ingredients mapped to this reference data, thereby providing a personalized skincare solution. FIG. 1 is a diagram illustrating the functional configuration of a selfie video-based artificial intelligence skincare service platform device (100) according to one embodiment of the present invention. FIG. 2 is a drawing illustrating an artificial intelligence skincare service platform device (100) based on a selfie video of FIG. 1. FIG. 3 is a diagram illustrating the system configuration of the selfie video-based artificial intelligence skincare service platform device (100) of FIG. 1. FIG. 4 is a flowchart explaining the operation of the selfie video-based artificial intelligence skincare service platform device (100) of FIG. 1. The description of the present invention is merely an example for structural or functional explanation, and therefore the scope of the present invention should not be interpreted as being limited by the examples described in the text. That is, since the examples are subject to various modifications and may take various forms, the scope of the present invention should be understood to include equivalents capable of realizing the technical concept. Furthermore, the objectives or effects presented in the present invention do not imply that a specific example must include all of them or only such effects; therefore, the scope of the present invention should not be understood as being limited by them. Meanwhile, the meaning of the terms described in this application should be understood as follows. Terms such as "first," "second," etc., are intended to distinguish one component from another, and the scope of rights shall not be limited by these terms. For example, the first component may be named the second component, and similarly, the second component may be named the first component. When it is stated that one component is "connected" to another component, it should be understood that it may be directly connected to that other component, or that there may be other components in between. Conversely, when it is stated that