WO2012060537A2 - 얼굴 및 스타일 인식 기반의 추천 시스템 및 그 방법 - Google Patents
얼굴 및 스타일 인식 기반의 추천 시스템 및 그 방법 Download PDFInfo
- Publication number
- WO2012060537A2 WO2012060537A2 PCT/KR2011/005210 KR2011005210W WO2012060537A2 WO 2012060537 A2 WO2012060537 A2 WO 2012060537A2 KR 2011005210 W KR2011005210 W KR 2011005210W WO 2012060537 A2 WO2012060537 A2 WO 2012060537A2
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- style
- information
- recommendation
- face
- characteristic
- Prior art date
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
- G06V40/168—Feature extraction; Face representation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
Definitions
- the present invention relates to a recommendation system and method based on face and style recognition, and in particular, extracts facial feature information and style feature information from a user image, and extracts facial feature information and style feature information from the extracted facial feature information and style feature information.
- the user's face is recommended by searching for and recommending recommendation style information (for example, hair style, makeup style, product information, etc.) matching the recognized facial characteristic and style characteristic from a template table of recommended styles for each characteristic.
- recommendation style information for example, hair style, makeup style, product information, etc.
- a user not only performs voice communication while carrying a mobile phone, but also uses a wireless Internet technology to wirelessly access the Internet to receive multimedia data services such as text, images, voice, or video.
- multimedia data services such as text, images, voice, or video.
- Additional features offered by mobile phones include music players, short message services, wireless messengers, mobile banking, fingerprint recognition for user authentication, and camera functions.
- face recognition technology is mounted on smartphones along with the craze of smartphones.
- Application technology using face recognition technology is expected to spread widely.
- Face recognition technology is a kind of bio-recognition technology and, unlike contact iris and fingerprint recognition, has been applied to various devices as a recognition technology that provides non-contact user convenience.
- a virtual experience service that allows a user to experience clothes, hairstyles, product information, and the like suitable for the user before visiting a store has been developed.
- the user checks the size, color, etc. in advance on the shopping mall site of the product to check whether the product is suitable for the user.
- the user can virtually experience the size or color of the product.
- the conventional virtual experience service may insert a virtual image of clothes or hair selected by a user into a real image, and provide a user with a real image into which the virtual image is inserted. This allows users to compare different clothes. This feature saves users time.
- the user can select a number of virtual styles one by one to see if they match their size or preferences, but since so many styles are selected one by one, it takes a lot of time or effort to find a style suitable for oneself.
- the conventional virtual experience service has a problem that it is difficult to find a style or product information suitable for itself as the styles that can be compared increase.
- the present invention was devised to solve the above problems, and extracts facial feature information and style feature information from a user image, and recognizes facial feature and style feature from the extracted facial feature information and style feature information. Matches the recommended style information (e.g., hair style, makeup style, product information, etc.) that matches the facial characteristics and style characteristics that are matched with the user's face and style by searching and recommending them in the recommended style table by characteristics pre-templated. It is an object of the present invention to provide a recommendation system based on face and style recognition, which can quickly and easily recommend recommended style information.
- the recommended style information e.g., hair style, makeup style, product information, etc.
- a user terminal for transmitting a user image through a communication network, or extracts the facial feature information and style feature information from the user image and transmits through the communication network; And generating recommendation style tables by template recommendation style information matching the facial characteristics and style characteristics, and recognizing the facial characteristics and style characteristics from the user image or the facial characteristic information and the style characteristic information transmitted from the user terminal. And a recommendation device that searches for recommendation style information matching the recognized face and style characteristics from the generated recommendation style table and transmits the recommended style information to the user terminal.
- the apparatus extracts facial feature information from the user image transmitted from the user terminal and recognizes the facial feature using the extracted facial feature information, or the facial feature information transmitted from the user terminal.
- a face recognition unit for recognizing face characteristics using the apparatus A style recognition unit extracting style feature information from the transmitted user image and recognizing a style feature using the extracted style feature information, or recognizing a style feature using the style feature information transmitted from the user terminal;
- a recommendation unit which searches for recommendation style information matching the recognized face characteristics and style characteristics from among recommendation style tables in which face characteristics and style characteristic recommendations are templated, and transmits them to the user terminal.
- the method according to the third aspect of the present invention information extraction step of extracting the facial feature information and style feature information from the user image; A face recognition step of recognizing a face characteristic using the extracted face feature information; A style recognition step of recognizing a style characteristic by using the extracted style feature information; And a style recommendation step of searching for recommendation style information matching the recognized characteristic and style characteristic from the recommendation style table in which the characteristic recommendation style information is templated and transmitting the same to the user terminal.
- the present invention extracts facial feature information and style feature information from a user image, recognizes a facial feature and a style feature from the extracted facial feature information and style feature information, and then recommends a style that matches the recognized feature and style feature.
- the present invention searches for and recommends the hairstyle information matching the recognized facial characteristics and style characteristics from among the hair style information for each of the previously learned facial characteristics, thereby quickly / easily searching for a hairstyle that best matches the user's face. There is an effect that can be recommended.
- the present invention further recognizes facial characteristics and style characteristics through not only facial feature points, forehead lengths and hair lengths extracted from user images, but also age and gender related to hair recommendation, and user's hairstyle preferences. It can help you recommend a suitable hairstyle.
- the present invention can easily construct a database frame for the recommendation style information by templated the recommendation style results recommended through the user image with the recommendation style information for each characteristic, more accurate based on the product recommendation results of other users It has the effect of recommending recommended style information.
- the present invention reflects not only facial feature point information extracted from the user image, but also style information related to the product recommendation and the user's product style preference in the product recommendation process, thereby making it possible to recommend a more suitable product style to the user. There is.
- FIG. 1 is a block diagram of an embodiment of a recommendation system based on face and style recognition according to the present invention
- FIG. 2 is a diagram illustrating an embodiment of a templated process of recommending style information and a product recommendation process according to the present invention
- FIG. 3 is a view illustrating an embodiment of a facial feature and style feature recognition process in the recommended device according to the present invention
- FIG. 4 is a diagram illustrating an embodiment of a hairstyle recommendation process according to the present invention.
- FIG. 5 is a flowchart of a first embodiment of a product recommendation method based on face and style recognition according to the present invention
- FIG. 6 is a flowchart of a second embodiment of a product recommendation method based on face and style recognition according to the present invention.
- FIG. 1 is a block diagram of an embodiment of a recommendation system based on face and style recognition according to the present invention.
- the recommendation system 10 includes a user terminal 101 and a recommendation apparatus 100.
- the recommendation device 100 is a templateization unit 110, face recognition unit 120, style recognition unit 130, recommendation unit 140, face DB 150, style DB 160, hair DB (170) ), The make-up DB 180 and the product DB (190).
- the user terminal 101 transmits the user image through a communication network, or facial feature information (eg, facial feature point information, skin color, wrinkle information, mouth shape, eye shape, brow, nose size, and forehead width, etc.) in the user image; Style feature information (eg, color information, clothing pattern information, season information, weather information, time information, etc.) is extracted and transmitted through a communication network.
- facial feature information eg, facial feature point information, skin color, wrinkle information, mouth shape, eye shape, brow, nose size, and forehead width, etc.
- Style feature information eg, color information, clothing pattern information, season information, weather information, time information, etc.
- the user terminal 101 transmits the user image to the recommendation apparatus 100 through a communication network.
- the user terminal 101 may be a computer, a mobile phone, or a smartphone equipped with an image capturing module, but is not limited thereto.
- the user terminal 101 acquires a user image by capturing an image of the user using the provided image capturing module.
- the image capturing module may be a camera or a webcam connected to an external control device such as a computer, or a camera embedded in a personal portable terminal.
- the user terminal 101 detects a face region of the user from an actual image acquired through the image capturing module, and extracts facial feature information from the detected face region.
- the user terminal 101 detects a user style region excluding the face region of the user from the actual image, and extracts style feature information from the detected user style region.
- the user terminal 101 transmits the extracted facial feature information and style feature information to the recommendation apparatus 100 through a communication network.
- the facial feature information includes facial feature point information on main parts of the face such as eyes, nose, mouth, and outline, the length of the forehead, and the head length from the forehead to the head.
- the facial feature information may include skin color, wrinkle information, mouth shape, eye shape, eyebrow shape, brow, nose shape, and the like.
- the style feature information may include color information, clothing pattern information, season information, weather information, indoor / outdoor information, and time information.
- the user terminal 101 may reduce or enlarge the actual image according to a preset face region size before detecting the face region or the user style region. The process of reducing and enlarging the actual image helps the user terminal 101 to accurately detect the face area and to detect the facial feature points thereafter.
- the recommendation apparatus 100 As a first embodiment of the recommendation apparatus 100, the recommendation apparatus 100 generates a recommendation style table by templateting recommendation style information for each characteristic through previously collected or simulated face feature information and style feature information.
- the recommendation apparatus 100 receives a user image from the user terminal 101, and extracts facial feature information and style feature information from the received user image. Subsequently, the recommendation apparatus 100 recognizes the facial feature and the style feature by using the extracted facial feature information and the style feature information.
- the facial feature information and the style feature information other than the user image are received from the user terminal 101, and the facial feature and style feature are received from the received facial feature information and the style feature information. Recognize.
- the recommendation apparatus 100 of the first and second exemplary embodiments searches for recommendation style information on a feature that matches the recognized facial feature and style feature in the recommendation style table.
- the recommendation device 100 transmits the found recommendation style information to the user terminal 101.
- the recommended style information includes at least one of hair style information, makeup style information, and recommended product information.
- the templater 110 generates a recommendation style table by templated recommendation style information for each characteristic by analyzing previously collected or simulated facial feature information and style feature information and corresponding recommendation style information.
- the templater 110 stores the templated recommendation style information for each characteristic in the DB among the hair DB 170, the makeup DB 180, and the product DB 190.
- the templater 110 matches the recognized facial feature and style feature information with the recommended style information retrieved by the recommender 140.
- the templater 110 templates the matching result into new style recommendation style information and stores the matching result in the corresponding DB among the hair DB 170, the makeup DB 180, and the product DB 190.
- new recommendation style information may be templated and stored in the hair DB 170, the makeup DB 180, and the product DB 190.
- the face recognition unit 120 extracts face feature information from the user image transmitted from the user terminal 101, and recognizes the face feature using the extracted face feature information.
- the face recognition unit 120 extracts facial feature information including facial feature point information, skin color, wrinkle information, nose size, forehead width, and the like from the user image transmitted from the user terminal 101.
- the face recognizing unit 120 recognizes facial characteristics by using facial feature information including extracted facial feature point information, skin color, wrinkle information, nose size, and forehead width.
- the facial recognition unit 120 may recognize the gender and age of the user by dividing the male / female and the teenager, 20s, 40s and the like.
- the face recognition unit 120 recognizes face characteristics by using a matching result between face feature information and face characteristics stored in the face DB 150.
- the facial features may include gender and age for style recommendation, and the overall facial features are further included.
- the face recognition unit 120 stores the facial feature information extracted from the user image and the recognized facial feature in the face DB 150.
- the style recognizing unit 130 extracts style feature information from the user image transmitted from the user terminal 101, and recognizes the style feature using the extracted style feature information.
- the style recognizing unit 130 extracts style feature information including color information, clothing pattern information, season information, weather information, indoor / outdoor information, and time information from the user image transmitted from the user terminal 101. That is, the style recognizing unit 130 recognizes a style characteristic by using style feature information including extracted color information, clothing pattern information, season information, weather information, indoor / outdoor information, and time information.
- the style recognition unit 130 may recognize the cool suit style from the style characteristic information by dividing into a beige color, a suit style, summer, sunny, outdoor and afternoon time.
- the style recognizing unit 130 recognizes the style characteristic by using a matching result between the style characteristic information and the style characteristic stored in the face DB 150.
- the style characteristic may include color information for clothing style recommendation, clothing pattern information, season information, weather information, indoor / outdoor information, and time information.
- the style recognizer 130 stores the style feature information and the recognized style feature extracted from the user image in the style DB 160.
- the recommendation unit 140 searches for recommendation style information on a feature matching the face and style characteristics recognized by the face recognition unit 120 in the recommendation style table.
- the recommender 140 may receive a style preference from the user terminal 101, and search for recommended style information matching the received style preference, face, and style characteristics.
- the recommendation unit 140 transmits the retrieved recommendation style information to the user terminal 101.
- the recommendation unit 140 may classify the searched recommendation style information according to a matching ratio with the characteristic and transmit the classified recommendation style information to the user terminal 101. For example, when there are a plurality of styles having a matching ratio greater than or equal to a specific ratio, the recommendation unit 140 may display and transmit the matching ratio for each recommendation style information.
- the user terminal 101 extracts facial feature information and style feature information from an actual image, recognizes the facial feature and style feature from the extracted feature information, and recognizes the feature.
- a series of processes for retrieving recommended style information matching face and style characteristics may be performed by oneself.
- the user terminal 101 includes a memory, a face recognizer, a style recognizer, and a recommender.
- the memory stores a recommendation style table in which recommendation style information matching the facial characteristic and the style characteristic is templated.
- the face recognizer includes a photographing module to capture a user and extract facial feature information from the captured user image.
- the face recognizer recognizes facial features using the extracted facial feature information.
- the style recognizer extracts style feature information from the captured user image and recognizes the style feature by using the extracted style feature information.
- the recommender searches for recommendation style information matching the face characteristics and style characteristics recognized by the face recognizer and the style recognizer from the recommended style table in which the face characteristics and the style recommendation information for each style characteristic are stored in the memory and provide them to the user. Can be.
- FIG. 2 is a diagram illustrating an embodiment of a template process and style recommendation process of recommendation style information according to the present invention.
- the style recommendation process in the recommendation apparatus 100 is largely a face recognition process 210, a style recognition process 220, a templated process 230 of recommended style information for each characteristic, and a recommended style search.
- Process 240 the style recommendation process in the recommendation apparatus 100 is largely a face recognition process 210, a style recognition process 220, a templated process 230 of recommended style information for each characteristic, and a recommended style search.
- the recommendation apparatus 100 performs a face recognition process 210, a style recognition process 220, and a template process 230 of recommendation style information for each characteristic.
- the recommendation apparatus 100 For the face recognition process 210 and the style recognition process 220, the recommendation apparatus 100 detects the face region 202 in the user image 201 transmitted from the user terminal 101, and detects the detected face region. In 202, face feature information is extracted. Subsequently, the recommendation apparatus 100 may recognize a gender and an age range from the extracted facial feature information. In addition, the recommendation apparatus 100 may extract style feature information from the user image 201 except for the face region 202, and recognize the style characteristic of the user from the extracted style feature information. The facial feature information and the facial feature, and the style feature information and the style feature are stored in the face DB 150 and the style DB 160, respectively.
- the recommendation apparatus 100 For the process of template 230 of recommendation style information for each feature, the recommendation apparatus 100 generates a recommendation style table using recommended style information matching the recognized face feature and style feature and stores the recommendation style table in a corresponding DB.
- the recommendation apparatus 100 After the process of template 230 of recommendation style information for each characteristic 230, the recommendation apparatus 100 performs a face recognition process 210 and a style recognition process 220 using the inputted new user image 203 and the face region 204. Perform facial recognition and style characteristics.
- the recommendation apparatus 100 searches for recommendation style information in the recommendation style table based on the recognized face and style characteristics.
- the recommendation apparatus 100 may search for recommendation style information matching the face characteristic and the style characteristic among the styles 1, 2, and 3 included in the recommendation style table stored in the product DB 190.
- the recommendation apparatus 100 may receive the recommendation style information by requesting the recommendation style information from an external style search mall.
- the user terminal 101 may receive a style preference from the user and transmit the style preference to the recommendation apparatus 100 to request a style recommendation.
- the purchase form of the individual customer may be reflected in the search process 240 of the recommendation style information.
- FIG. 3 is a diagram illustrating an embodiment of a facial feature and style feature recognition process in the recommended device according to the present invention.
- the face recognition unit 120 may analyze a gender (male and female) and an age group through the face recognition process 210. As illustrated in FIG. 3, the face recognition unit 120 may extract face feature information for each of the plurality of users from the user image 203, and analyze the gender and age group of each user from the extracted face feature information. . As a result, the face recognition unit 120 may recognize a gender and age group of each user as a male, a user 1-10 years old, a female, a 31-40 year old user, a female, 11-20 year old user, or the like.
- the style recognizing unit 130 may extract style feature information of each user in an area excluding the face area 204 in the user image 203, thereby recognizing style characteristics.
- the style recognition unit 130 is a male and a user who is 1 to 10 years old, the color is light blue, the clothing pattern is a t-shirt, the season is autumn, the weather is sunny, and the characteristic is 2 pm
- the information may be extracted, and the style characteristic of the user 1-10 years old may be extracted from the extracted style characteristic information.
- FIG. 4 is a diagram illustrating an embodiment of a hairstyle recommendation process according to the present invention.
- the user terminal 101 may include facial feature point information 411, a forehead length 412, and a forehead to a head in a user image 410 captured or acquired from an external image capturing module. Head length 413 is extracted.
- the facial feature point information 411, the length of the forehead 412, and the length from the forehead to the head 413 are necessary information for hair recommendation, and may further include gender and age information of the user.
- the user terminal 101 transmits the extracted facial feature point information 411, the length 412 of the forehead, and the hair length 413 from the forehead to the head to the recommendation apparatus 100 to request a hairstyle recommendation.
- the user terminal 101 may receive a hair style preference from the user and transmit the hair style preference to the recommendation apparatus 100 to request hair recommendation.
- the recommendation apparatus 100 searches for the hairstyle information matching the facial characteristics recognized by the face recognition unit 120 through the recommendation unit 140, and retrieves the retrieved hairstyle information 420 through the communication network. Recommend hairstyles by sending to 101.
- the hairstyle information 420 may be a hairstyle image in which only a hairstyle is expressed, and may be a virtual hairstyle experience image in which a hairstyle is inserted into a user image.
- FIG. 5 is a flowchart of a first embodiment of a recommendation method based on face and style recognition according to the present invention.
- the templater 110 generates a recommendation style table by analyzing facial feature information and style feature information and corresponding recommendation style information, and template recommendation style information for each characteristic (S502).
- the facial feature information, the style feature information, and the corresponding recommendation style information are templated, generated as a recommendation style table, and stored in the hair DB 170, the makeup DB 180, and the product DB 190, which are corresponding DBs.
- the face recognition unit 120 and the style recognition unit 130 extract face facial feature information and style feature information from the user image transmitted from the user terminal 101, respectively (S504).
- the face recognition unit 120 extracts face feature information including face feature point information, skin color, wrinkle information, and the like from the user image transmitted from the user terminal 101.
- the style recognition unit 130 extracts style feature information including color information, clothing pattern information, season information, weather information, and the like from the user image.
- the face recognition unit 120 recognizes face characteristics using the extracted face feature information (S506).
- the face recognizing unit 120 recognizes style characteristics using facial feature point information, forehead length, and forehead length.
- the face recognition unit 120 may recognize a gender and an age range from the extracted face feature information.
- the style recognizing unit 130 recognizes the style characteristic by using the extracted style characteristic information (S508).
- the style recognizing unit 130 may recognize style characteristics through the extracted color information, clothing pattern information, season information, weather information, and the like.
- the recommendation unit 140 recommends style information for characteristics matching the facial characteristics and the style characteristics recognized by the face recognition unit 120 and the style recognition unit 130, and recommends by characteristics generated in step S502.
- the style table is searched for (S510).
- the recommended style information includes at least one of hair style information, makeup style information, and recommended product information.
- the recommender 140 may receive a product style preference from the user terminal 101 and search for recommendation style information matching the received product style preference and characteristics.
- the recommendation unit 140 may classify the searched recommendation style information according to a matching ratio with a characteristic.
- the recommendation unit 140 transmits the found recommendation style information to the user terminal 101 (S512).
- the templater 110 matches the feature recognized by the face recognition unit 120 and the style feature unit 130 with the recommended style information retrieved by the recommender 140, and the matching result is a new feature. It can be templated with star recommendation style information.
- FIG. 6 is a flowchart of a second embodiment of a recommendation method based on face and style recognition according to the present invention.
- the templater 110 analyzes the recommended style information matching the face characteristic and the style characteristic to template the recommended style information for each characteristic (S602).
- the recommendation style information matching the facial feature information and the style feature information may be information collected in advance or simulated and stored in the product DB 190.
- the user terminal 101 extracts facial feature information including facial feature point information, skin color, and wrinkle information from the user image photographed through the image capturing module, and transmits the feature information to the recommendation apparatus 100.
- the user terminal 101 extracts style feature information including color information, clothing pattern information, season information, and the like from the user image and transmits it to the recommendation apparatus 100.
- the face recognition unit 120 receives face feature information and style feature information extracted from the user terminal 101 (S604).
- the face recognition unit 120 recognizes face characteristics and style characteristics by using face characteristic information transmitted from the user terminal 101 (S606).
- the face recognition unit 120 recognizes facial features using facial feature point information, forehead length, and forehead to head length.
- the face recognition unit 120 may recognize the gender and the age group of the user.
- the face recognition unit 130 recognizes the style characteristic by using the style characteristic information transmitted from the user terminal 101 (S608).
- the style recognizing unit 130 recognizes the style characteristic from the style characteristic information including color information, clothing pattern information, season information, and the like (S608).
- the recommendation unit 140 includes recommendation style information that matches the facial characteristics recognized by the face recognition unit 120 and the style characteristics recognized by the style recognition unit 130, and recommendation styles in which recommended style information for each characteristic is templated. Search in the table (S610).
- the recommendation unit 140 transmits the found recommendation style information to the user terminal 101 (S612).
- the templater 110 may match the facial feature and style feature with the recommendation style information retrieved from the recommender 140, and template the matching result into new recommendation style information for each feature.
- the user terminal 101 when the user terminal 101 independently performs a service without receiving a network-based service, the user terminal 101 includes a face recognizer, a style recognizer, and a recommender.
- the recommendation style table with matching recommendation style information is stored in advance.
- the user terminal 101 photographs the user through the photographing module provided.
- the user terminal 101 extracts facial feature information from the captured user image. In addition, the user terminal 101 recognizes a facial feature by using the extracted facial feature information.
- the user terminal 101 extracts style feature information from the captured user image and recognizes the style feature by using the extracted style feature information.
- the user terminal 101 searches for recommendation style information matching the face characteristics and style characteristics recognized by the face recognizer and the style recognizer from the recommended style table in which the face characteristics and the recommended style information for each style characteristic are stored in the memory. Can be provided to the user.
- the present invention extracts facial feature information and style feature information from a user image, recognizes a facial feature and a style feature from the extracted facial feature information and style feature information, and then recommends matching the recognized facial feature and style feature.
- Quickly and easily recommend recommended style information that best matches the user's face and style by searching for and recommending style information (e.g., hair style, makeup style, product information, etc.) in the recommended style table by characteristic pre-templated. I can do it.
- style information e.g., hair style, makeup style, product information, etc.
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Strategic Management (AREA)
- Accounting & Taxation (AREA)
- Development Economics (AREA)
- Finance (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Entrepreneurship & Innovation (AREA)
- Game Theory and Decision Science (AREA)
- General Business, Economics & Management (AREA)
- Marketing (AREA)
- Economics (AREA)
- Health & Medical Sciences (AREA)
- Oral & Maxillofacial Surgery (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Health & Medical Sciences (AREA)
- Human Computer Interaction (AREA)
- Multimedia (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
Description
Claims (16)
- 사용자 영상을 통신망을 통해 전송하거나, 상기 사용자 영상에서 얼굴 특징 정보 및 스타일 특징 정보를 추출하여 통신망을 통해 전송하는 사용자 단말; 및얼굴 특성 및 스타일 특성과 매칭되는 추천 스타일 정보를 템플릿화하여 추천 스타일 테이블을 생성하고, 상기 사용자 단말로부터 전송된 사용자 영상 또는 얼굴 특징 정보 및 스타일 특징 정보로부터 얼굴 특성 및 스타일 특성을 인식하고, 상기 인식된 얼굴 특성 및 스타일 특성과 매칭되는 추천 스타일 정보를 상기 생성된 추천 스타일 테이블 중에서 검색하여 상기 사용자 단말로 전송하는 추천 장치를 포함하는 얼굴 및 스타일 인식 기반의 추천 시스템.
- 사용자 단말로부터 전송된 사용자 영상에서 얼굴 특징 정보를 추출하고 상기 추출된 얼굴 특징 정보를 이용하여 얼굴 특성을 인식하거나, 사용자 단말로부터 전송된 얼굴 특징 정보를 이용하여 얼굴 특성을 인식하는 얼굴 인식부;상기 전송된 사용자 영상에서 스타일 특징 정보를 추출하고 상기 추출된 스타일 특징 정보를 이용하여 스타일 특성을 인식하거나, 상기 사용자 단말로부터 전송된 스타일 특징 정보를 이용하여 스타일 특성을 인식하는 스타일 인식부; 및상기 인식된 얼굴 특성 및 스타일 특성과 매칭되는 추천 스타일 정보를 얼굴 특성 및 스타일 특성별 추천 스타일 정보가 템플릿화된 추천 스타일 테이블 중에서 검색하여 상기 사용자 단말로 전송하는 추천부를 포함하는 얼굴 및 스타일 인식 기반의 추천 장치.
- 제 2 항에 있어서,상기 추천부는,헤어 스타일 정보, 메이크업 스타일 정보 및 추천 상품 정보 중 적어도 하나를 상기 추천 스타일 정보에 포함시켜 상기 사용자 단말로 전송하는 것을 특징으로 하는 얼굴 및 스타일 인식 기반의 추천 장치.
- 제 2 항에 있어서,수집된 특성 및 스타일 특성과 매칭된 추천 스타일 정보를 분류하고, 상기 분류 결과에 따라 특성별 추천 스타일 정보를 템플릿화하여 상기 추천 스타일 테이블을 생성하는 템플릿화부를 더 포함하는 얼굴 및 스타일 인식 기반의 추천 장치.
- 제 2 항에 있어서,상기 얼굴 특징 정보 및 상기 인식된 얼굴 특성을 저장하는 얼굴 DB;상기 스타일 특징 정보 및 상기 인식된 스타일 특성을 저장하는 스타일 DB:상기 인식된 얼굴 특성 및 스타일 특성과 매칭된 헤어 스타일 정보를 저장하는 헤어 DB;상기 인식된 얼굴 특성 및 스타일 특성과 매칭된 메이크업 스타일 정보를 저장하는 메이크업 DB; 및상기 인식된 얼굴 특성 및 스타일 특성과 매칭된 추천 상품 정보를 저장하는 상품 DB를 더 포함하는 얼굴 및 스타일 인식 기반의 추천 장치.
- 제 2 항에 있어서,상기 얼굴 인식부는,상기 추출된 얼굴 특징 정보의 입 모양, 눈 모양, 코 모양, 미간, 피부 색상 및 주름 정보, 이마 넓이 중 적어도 하나로부터 상기 사용자의 성별 및 연령대를 상기 얼굴 특성으로 인식하는 얼굴 및 스타일 인식 기반의 추천 장치.
- 제 2 항에 있어서,상기 스타일 인식부는,상기 추출된 스타일 특징 정보의 의류 패턴 정보, 색상 정보, 계절 정보 및 날씨 정보 중 적어도 하나로부터 상기 사용자의 스타일 특성을 인식하는 얼굴 및 스타일 인식 기반의 추천 장치.
- 제 2 항에 있어서,상기 추천부는,상기 사용자 단말에 의해 상기 인식된 스타일 특성이 변경되거나 스타일 특성이 추가되는 경우, 상기 변경되거나 추가된 스타일 특성과 매칭되는 추천 스타일 정보를 재검색하여 상기 사용자 단말로 전송하는 얼굴 및 스타일 인식 기반의 추천 장치.
- 제 2 항에 있어서,상기 추천부는,상기 검색된 추천 스타일 정보가 복수인 경우, 상기 검색된 복수의 추천 스타일 정보를 상기 인식된 특성 및 스타일 특성과의 매칭 비율에 따라 우선 순위를 구분하여 상기 사용자 단말로 전송하는 얼굴 및 스타일 인식 기반의 추천 장치.
- 사용자 영상에서 얼굴 특징 정보 및 스타일 특징 정보를 추출하는 정보 추출 단계;상기 추출된 얼굴 특징 정보를 이용하여 얼굴 특성을 인식하는 얼굴 인식 단계;상기 추출된 스타일 특징 정보를 이용하여 스타일 특성을 인식하는 스타일 인식 단계; 및상기 인식된 특성 및 스타일 특성과 매칭되는 추천 스타일 정보를 특성별 추천 스타일 정보가 템플릿화된 추천 스타일 테이블 중에서 검색하여 사용자 단말로 전송하는 스타일 추천 단계를 포함하는 얼굴 및 스타일 인식 기반의 상품 추천 방법.
- 제 10 항에 있어서,상기 스타일 추천 단계는,헤어 스타일 정보, 메이크업 스타일 정보 및 추천 상품 정보 중 적어도 하나를 상기 추천 스타일 정보에 포함시켜 상기 사용자 단말로 전송하는 것을 특징으로 하는 얼굴 및 스타일 인식 기반의 추천 방법.
- 제 10 항에 있어서,수집된 특성 및 스타일 특성과 매칭된 추천 스타일 정보를 분류하고, 상기 분류 결과에 따라 특성별 추천 스타일 정보를 템플릿화하여 상기 추천 스타일 테이블을 생성하는 추천 상품 템플릿화 단계를 더 포함하는 얼굴 및 스타일 인식 기반의 상품 추천 방법.
- 제 10 항에 있어서,상기 얼굴 인식 단계는,상기 추출된 얼굴 특징 정보의 입 모양, 눈 모양, 코 모양, 미간, 피부 색상 및 주름 정보, 이마 넓이 중 적어도 하나로부터 상기 사용자의 성별 및 연령대를 상기 얼굴 특성으로 인식하는 얼굴 및 스타일 인식 기반의 상품 추천 방법.
- 제 10 항에 있어서,상기 스타일 인식 단계는,상기 추출된 스타일 특징 정보의 의류 패턴 정보, 색상 정보, 계절 정보 및 날씨 정보 중 적어도 하나로부터 상기 사용자의 스타일 특성을 인식하는 얼굴 및 스타일 인식 기반의 상품 추천 방법.
- 제 10 항에 있어서,상기 스타일 추천 단계는,상기 사용자 단말에 의해 상기 인식된 스타일 특성이 변경되거나 스타일 특성이 추가되는 경우, 상기 변경되거나 추가된 스타일 특성과 매칭되는 추천 스타일 정보를 재검색하여 상기 사용자 단말로 전송하는 얼굴 및 스타일 인식 기반의 상품 추천 방법.
- 제 10 항에 있어서,상기 스타일 추천 단계는,상기 검색된 추천 스타일 정보가 복수인 경우, 상기 검색된 복수의 추천 스타일 정보를 상기 인식된 특성 및 스타일 특성과의 매칭 비율에 따라 우선 순위를 구분하여 상기 사용자 단말로 전송하는 얼굴 및 스타일 인식 기반의 상품 추천 방법.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US13/813,003 US20130129210A1 (en) | 2010-11-02 | 2011-07-15 | Recommendation system based on the recognition of a face and style, and method thereof |
Applications Claiming Priority (4)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
KR1020100108442A KR20120046653A (ko) | 2010-11-02 | 2010-11-02 | 얼굴 및 스타일 인식 기반의 상품 추천 시스템 및 그 방법 |
KR10-2010-0108442 | 2010-11-02 | ||
KR1020100108441A KR20120046652A (ko) | 2010-11-02 | 2010-11-02 | 얼굴 인식 기반의 헤어 추천 시스템 및 그 방법 |
KR10-2010-0108441 | 2010-11-02 |
Publications (2)
Publication Number | Publication Date |
---|---|
WO2012060537A2 true WO2012060537A2 (ko) | 2012-05-10 |
WO2012060537A3 WO2012060537A3 (ko) | 2012-06-28 |
Family
ID=46024896
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/KR2011/005210 WO2012060537A2 (ko) | 2010-11-02 | 2011-07-15 | 얼굴 및 스타일 인식 기반의 추천 시스템 및 그 방법 |
Country Status (2)
Country | Link |
---|---|
US (1) | US20130129210A1 (ko) |
WO (1) | WO2012060537A2 (ko) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2014106213A1 (en) * | 2012-12-31 | 2014-07-03 | Agrawal Vandana | Style recommendation engine and method |
CN107545051A (zh) * | 2017-08-23 | 2018-01-05 | 武汉理工大学 | 基于图像处理的发型设计系统及方法 |
Families Citing this family (41)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20150052008A1 (en) * | 2013-08-16 | 2015-02-19 | iWeave International | Mobile Application For Hair Extensions |
CN103870821A (zh) * | 2014-04-10 | 2014-06-18 | 上海影火智能科技有限公司 | 一种虚拟试妆的方法和系统 |
US9760935B2 (en) * | 2014-05-20 | 2017-09-12 | Modiface Inc. | Method, system and computer program product for generating recommendations for products and treatments |
US10235388B2 (en) * | 2014-06-27 | 2019-03-19 | Ebay Inc. | Obtaining item listings relating to a look of image selected in a user interface |
KR102077260B1 (ko) | 2014-08-08 | 2020-02-13 | 삼성전자주식회사 | 확룔 모델에 기반한 신뢰도를 이용하여 얼굴을 인식하는 방법 및 장치 |
CN105741256B (zh) * | 2014-12-09 | 2020-08-04 | 富泰华工业(深圳)有限公司 | 电子设备及其刮须提示系统与方法 |
CN104866589B (zh) * | 2015-05-28 | 2018-06-15 | 北京京东尚科信息技术有限公司 | 数据报表的生成方法和装置 |
CN106354734B (zh) * | 2015-07-17 | 2019-06-11 | 阿里巴巴集团控股有限公司 | 提供业务对象信息的方法及装置 |
CN105204709B (zh) * | 2015-07-22 | 2019-10-18 | 维沃移动通信有限公司 | 主题切换的方法及装置 |
WO2017030255A1 (en) * | 2015-08-18 | 2017-02-23 | Samsung Electronics Co., Ltd. | Large format display apparatus and control method thereof |
US9811762B2 (en) * | 2015-09-22 | 2017-11-07 | Swati Shah | Clothing matching system and method |
US20170148076A1 (en) * | 2015-11-25 | 2017-05-25 | Electronics And Telecommunications Research Institute | Method for operating personal information brokerage apparatus and method for operating customized product production system using the same |
CN106887024B (zh) * | 2015-12-16 | 2019-09-17 | 腾讯科技(深圳)有限公司 | 照片的处理方法及处理系统 |
TWI626610B (zh) * | 2015-12-21 | 2018-06-11 | 財團法人工業技術研究院 | 訊息推播方法與訊息推播裝置 |
US9996981B1 (en) | 2016-03-07 | 2018-06-12 | Bao Tran | Augmented reality system |
US9460557B1 (en) | 2016-03-07 | 2016-10-04 | Bao Tran | Systems and methods for footwear fitting |
US10497014B2 (en) * | 2016-04-22 | 2019-12-03 | Inreality Limited | Retail store digital shelf for recommending products utilizing facial recognition in a peer to peer network |
CN106250541A (zh) * | 2016-08-09 | 2016-12-21 | 珠海市魅族科技有限公司 | 一种信息的推送方法及装置 |
CN109982616B (zh) * | 2016-09-27 | 2021-10-08 | 皇家飞利浦有限公司 | 用于支持至少一个用户执行个人护理活动的装置和方法 |
US11164195B2 (en) | 2017-02-14 | 2021-11-02 | International Business Machines Corporation | Increasing sales efficiency by identifying customers who are most likely to make a purchase |
US10052026B1 (en) | 2017-03-06 | 2018-08-21 | Bao Tran | Smart mirror |
CN107391599B (zh) * | 2017-06-30 | 2021-01-12 | 中原智慧城市设计研究院有限公司 | 基于风格特征的图像检索方法 |
EP3669318A1 (de) * | 2017-08-16 | 2020-06-24 | Henkel AG & Co. KGaA | Verfahren und vorrichtung zur computergestützten haarbehandlungsberatung |
CN108234591B (zh) * | 2017-09-21 | 2021-01-05 | 深圳市商汤科技有限公司 | 基于身份验证装置的内容数据推荐方法、装置和存储介质 |
CA3078645A1 (en) * | 2017-10-18 | 2019-04-25 | Inreality Limited | Expedite processing of facial recognition of people in a local network |
US10646022B2 (en) | 2017-12-21 | 2020-05-12 | Samsung Electronics Co. Ltd. | System and method for object modification using mixed reality |
EP3511893A1 (en) * | 2018-01-12 | 2019-07-17 | Koninklijke Philips N.V. | Hair style recommendation apparatus |
CN108133055A (zh) * | 2018-01-23 | 2018-06-08 | 京东方科技集团股份有限公司 | 智能服饰存储装置及基于其的存储、推荐方法与装置 |
CN112292709A (zh) * | 2018-05-16 | 2021-01-29 | 美康美环球有限公司 | 用于提供发型推荐的系统和方法 |
KR102530264B1 (ko) * | 2018-08-08 | 2023-05-09 | 삼성전자 주식회사 | 아바타에 대응하는 속성에 따른 아이템을 제공하는 방법 및 장치 |
CN109447895B (zh) * | 2018-09-03 | 2021-06-08 | 腾讯科技(武汉)有限公司 | 图片生成方法和装置、存储介质及电子装置 |
US10929915B2 (en) | 2018-09-29 | 2021-02-23 | Wipro Limited | Method and system for multi-modal input based platform for intent based product recommendations |
CN109544262A (zh) * | 2018-09-30 | 2019-03-29 | 百度在线网络技术(北京)有限公司 | 物品推荐方法、装置、电子设备、系统及可读存储介质 |
CN111325705A (zh) * | 2018-11-28 | 2020-06-23 | 北京京东尚科信息技术有限公司 | 图像处理方法、装置、设备及存储介质 |
US11253045B2 (en) | 2019-07-18 | 2022-02-22 | Perfect Mobile Corp. | Systems and methods for recommendation of makeup effects based on makeup trends and facial analysis |
CN110516099A (zh) | 2019-08-27 | 2019-11-29 | 北京百度网讯科技有限公司 | 图像处理方法和装置 |
US11257139B2 (en) | 2019-08-28 | 2022-02-22 | Bank Of America Corporation | Physical needs tool |
CN110933354B (zh) * | 2019-11-18 | 2023-09-01 | 深圳传音控股股份有限公司 | 一种可定制的多风格多媒体处理方法及其终端 |
US11587358B2 (en) | 2020-03-26 | 2023-02-21 | Panasonic Avionics Corporation | Managing content on in-flight entertainment platforms |
CN111611920A (zh) * | 2020-05-21 | 2020-09-01 | 杭州智珺智能科技有限公司 | 一种基于属性特征提取的ai人脸风格辨识方法 |
CN117036203B (zh) * | 2023-10-08 | 2024-01-26 | 杭州黑岩网络科技有限公司 | 一种智能绘图方法及系统 |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2007065146A (ja) * | 2005-08-30 | 2007-03-15 | Fujifilm Corp | 画像注文システム |
JP2007514332A (ja) * | 2003-09-08 | 2007-05-31 | カライズマン,ギョーラ | 移動通信装置を使用する紹介システムおよび方法 |
JP2009251832A (ja) * | 2008-04-03 | 2009-10-29 | Sony Ericsson Mobilecommunications Japan Inc | ユーザ相関図作成装置、ユーザ相関図作成方法、ユーザ相関図作成プログラム、及びユーザ相関図作成システム |
KR20100069395A (ko) * | 2008-12-16 | 2010-06-24 | 주식회사 케이티 | 얼굴 인식 기술을 이용하여 사용자에 따라 특화된 아이피티브이 콘텐츠를 추천하는 시스템 및 방법 |
Family Cites Families (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
AU1459801A (en) * | 1999-11-04 | 2001-05-14 | Stefano Soatto | System for selecting and designing eyeglass frames |
JP2005321986A (ja) * | 2004-05-07 | 2005-11-17 | Pioneer Electronic Corp | ヘアスタイル提案システム、ヘアスタイル提案方法、及びコンピュータプログラム |
US20070058858A1 (en) * | 2005-09-09 | 2007-03-15 | Michael Harville | Method and system for recommending a product based upon skin color estimated from an image |
US20070073799A1 (en) * | 2005-09-29 | 2007-03-29 | Conopco, Inc., D/B/A Unilever | Adaptive user profiling on mobile devices |
JP5586436B2 (ja) * | 2009-12-03 | 2014-09-10 | パナソニック インテレクチュアル プロパティ コーポレーション オブ アメリカ | 生活スタイル収集装置、ユーザインターフェース装置及び生活スタイル収集方法 |
US9002700B2 (en) * | 2010-05-13 | 2015-04-07 | Grammarly, Inc. | Systems and methods for advanced grammar checking |
-
2011
- 2011-07-15 WO PCT/KR2011/005210 patent/WO2012060537A2/ko active Application Filing
- 2011-07-15 US US13/813,003 patent/US20130129210A1/en not_active Abandoned
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2007514332A (ja) * | 2003-09-08 | 2007-05-31 | カライズマン,ギョーラ | 移動通信装置を使用する紹介システムおよび方法 |
JP2007065146A (ja) * | 2005-08-30 | 2007-03-15 | Fujifilm Corp | 画像注文システム |
JP2009251832A (ja) * | 2008-04-03 | 2009-10-29 | Sony Ericsson Mobilecommunications Japan Inc | ユーザ相関図作成装置、ユーザ相関図作成方法、ユーザ相関図作成プログラム、及びユーザ相関図作成システム |
KR20100069395A (ko) * | 2008-12-16 | 2010-06-24 | 주식회사 케이티 | 얼굴 인식 기술을 이용하여 사용자에 따라 특화된 아이피티브이 콘텐츠를 추천하는 시스템 및 방법 |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2014106213A1 (en) * | 2012-12-31 | 2014-07-03 | Agrawal Vandana | Style recommendation engine and method |
CN107545051A (zh) * | 2017-08-23 | 2018-01-05 | 武汉理工大学 | 基于图像处理的发型设计系统及方法 |
Also Published As
Publication number | Publication date |
---|---|
WO2012060537A3 (ko) | 2012-06-28 |
US20130129210A1 (en) | 2013-05-23 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
WO2012060537A2 (ko) | 얼굴 및 스타일 인식 기반의 추천 시스템 및 그 방법 | |
KR20120046652A (ko) | 얼굴 인식 기반의 헤어 추천 시스템 및 그 방법 | |
KR20120046653A (ko) | 얼굴 및 스타일 인식 기반의 상품 추천 시스템 및 그 방법 | |
WO2019164266A1 (en) | Electronic device for generating image including 3d avatar reflecting face motion through 3d avatar corresponding to face and method of operating same | |
CN108234591B (zh) | 基于身份验证装置的内容数据推荐方法、装置和存储介质 | |
WO2016180285A1 (zh) | 智能眼镜 | |
WO2011081379A2 (en) | Display device and control method thereof | |
US9443307B2 (en) | Processing of images of a subject individual | |
WO2012020927A1 (ko) | 통합 영상 검색 시스템 및 그 서비스 방법 | |
WO2016072714A1 (en) | Electronic device and method for providing filter in electronic device | |
WO2019033571A1 (zh) | 面部特征点检测方法、装置及存储介质 | |
CN103942705A (zh) | 一种基于人脸识别的广告分类匹配推送方法及系统 | |
CN103412877A (zh) | 图片传递方法及装置 | |
WO2018012928A1 (ko) | 얼굴 인식을 이용한 사용자 인증 방법 및 그 장치 | |
WO2020032487A1 (ko) | 우선 순위에 기반하여 상품과 관련된 정보를 제공하는 방법 및 그 전자 장치 | |
WO2014088125A1 (ko) | 영상 촬영 장치 및 그 방법 | |
WO2015102126A1 (ko) | 얼굴 인식 기술을 이용한 전자 앨범 관리 방법 및 시스템 | |
WO2017142361A1 (ko) | 스타일 특징을 이용한 상품 추천 방법 | |
EP3366038A1 (en) | Method for synthesizing image and an electronic device using the same | |
KR20150007403A (ko) | 인물 인식 방법, 그리고 인물 정보 데이터 운용 방법 및 장치 | |
WO2018043923A1 (ko) | 디스플레이장치 및 그 제어방법 | |
TW201521415A (zh) | 通訊裝置及其來電管理方法 | |
CN103414821A (zh) | 图片传递方法及装置 | |
KR20120076492A (ko) | 사용자 상황 인식 기반의 가상 체험 시스템 및 그 방법 | |
WO2020145468A1 (ko) | 두발 상태 모니터링 시스템 및 이를 수행하기 위한 컴퓨팅 장치 |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 11838142 Country of ref document: EP Kind code of ref document: A2 |
|
WWE | Wipo information: entry into national phase |
Ref document number: 13813003 Country of ref document: US |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
32PN | Ep: public notification in the ep bulletin as address of the adressee cannot be established |
Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205A DATED 06/SEP/2013) |
|
122 | Ep: pct application non-entry in european phase |
Ref document number: 11838142 Country of ref document: EP Kind code of ref document: A2 |