EA202092529A1 - METHOD FOR TRAINING A NEURAL NETWORK FOR HUMAN FACE RECOGNITION - Google Patents

METHOD FOR TRAINING A NEURAL NETWORK FOR HUMAN FACE RECOGNITION

Info

Publication number
EA202092529A1
EA202092529A1 EA202092529A EA202092529A EA202092529A1 EA 202092529 A1 EA202092529 A1 EA 202092529A1 EA 202092529 A EA202092529 A EA 202092529A EA 202092529 A EA202092529 A EA 202092529A EA 202092529 A1 EA202092529 A1 EA 202092529A1
Authority
EA
Eurasian Patent Office
Prior art keywords
mini
double
image
neural network
training
Prior art date
Application number
EA202092529A
Other languages
Russian (ru)
Inventor
Евгений Алексеевич СМИРНОВ
Original Assignee
Общество с ограниченной ответственностью "ЦРТ-инновации"
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Общество с ограниченной ответственностью "ЦРТ-инновации" filed Critical Общество с ограниченной ответственностью "ЦРТ-инновации"
Publication of EA202092529A1 publication Critical patent/EA202092529A1/en

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N20/00—Machine learning
    • G06N20/10—Machine learning using kernel methods, e.g. support vector machines [SVM]
    • 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
    • G06V40/161—Detection; Localisation; Normalisation

Abstract

Изобретение относится к области лицевой биометрии, в частности к задаче обучения нейронных сетей для распознавания лиц. Предложен способ обучения нейронных сетей, согласно которому обеспечивают наличие базы данных с изображениями лиц людей и обеспечивают наличие списка двойников. После этого формируют мини-пакет из изображений лиц людей путём сначала включения в него набора изображений лиц людей из базы данных, а затем добавления для каждого человека, по меньшей мере одно изображение которого включено в мини-пакет, по меньшей мере одного изображения его двойника из списка двойников, при наличии двойника и если изображение этого двойника ещё не добавлено в мини-пакет, а при отсутствии двойника или если изображение двойника уже включено в мини-пакет, добавления по меньшей мере одного изображения другого человека из базы данных. Далее подают изображения лиц людей из мини-пакета на вход нейронной сети. Формируют верификационный и идентификационный обучающие сигналы с использованием результатов, полученных на выходе нейронной сети. После этого обучают нейронную сеть с использованием верификационного и идентификационного обучающего сигнала. При этом ставят в соответствие каждому человеку в качестве двойника другого человека с использованием указанных результатов с обновлением списка двойников при получении пары двойников, отсутствующей в списке двойников. Повторяют указанные операции начиная с формирования мини-пакета.The invention relates to the field of facial biometrics, in particular to the problem of training neural networks for face recognition. A method for training neural networks is proposed, according to which a database with images of human faces is provided and a list of twins is provided. After that, a mini-package is formed from images of people's faces by first including a set of images of people's faces from the database into it, and then adding, for each person, at least one image of which is included in the mini-package, at least one image of his double from list of doubles, if there is a double and if the image of this double has not yet been added to the mini-package, and if there is no double or if the image of the double is already included in the mini-package, add at least one image of another person from the database. Next, images of people's faces from the mini-packet are fed to the input of the neural network. Verification and identification training signals are generated using the results obtained at the output of the neural network. After that, the neural network is trained using the verification and identification training signal. At the same time, each person is matched as a double of another person using the indicated results, with the list of doubles being updated when a pair of doubles is obtained that is not in the list of doubles. These operations are repeated starting from the formation of a mini-packet.

EA202092529A 2018-04-23 2018-04-23 METHOD FOR TRAINING A NEURAL NETWORK FOR HUMAN FACE RECOGNITION EA202092529A1 (en)

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/RU2018/000259 WO2019209131A1 (en) 2018-04-23 2018-04-23 Method of training a neural network for human facial recognition

Publications (1)

Publication Number Publication Date
EA202092529A1 true EA202092529A1 (en) 2021-02-02

Family

ID=68295642

Family Applications (1)

Application Number Title Priority Date Filing Date
EA202092529A EA202092529A1 (en) 2018-04-23 2018-04-23 METHOD FOR TRAINING A NEURAL NETWORK FOR HUMAN FACE RECOGNITION

Country Status (3)

Country Link
KR (1) KR20210033940A (en)
EA (1) EA202092529A1 (en)
WO (1) WO2019209131A1 (en)

Families Citing this family (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111126563B (en) * 2019-11-25 2023-09-29 中国科学院计算技术研究所 Target identification method and system based on space-time data of twin network
CN111325736B (en) * 2020-02-27 2024-02-27 成都航空职业技术学院 Eye differential image-based sight angle estimation method
CN113065645B (en) * 2021-04-30 2024-04-09 华为技术有限公司 Twin attention network, image processing method and device
CN114448664B (en) * 2021-12-22 2024-01-02 深信服科技股份有限公司 Method and device for identifying phishing webpage, computer equipment and storage medium
CN117273747B (en) * 2023-09-28 2024-04-19 广州佳新智能科技有限公司 Payment method, device, storage medium and equipment based on face image recognition

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP4591215B2 (en) * 2005-06-07 2010-12-01 株式会社日立製作所 Facial image database creation method and apparatus
US10860887B2 (en) * 2015-11-16 2020-12-08 Samsung Electronics Co., Ltd. Method and apparatus for recognizing object, and method and apparatus for training recognition model
CN106503669B (en) * 2016-11-02 2019-12-10 重庆中科云丛科技有限公司 Training and recognition method and system based on multitask deep learning network

Also Published As

Publication number Publication date
KR20210033940A (en) 2021-03-29
WO2019209131A1 (en) 2019-10-31

Similar Documents

Publication Publication Date Title
EA202092529A1 (en) METHOD FOR TRAINING A NEURAL NETWORK FOR HUMAN FACE RECOGNITION
EP3511942A3 (en) Cross-domain image analysis and cross-domain image synthesis using deep image-to-image networks and adversarial networks
CN106156757A (en) Face identification method and face identification system in conjunction with In vivo detection technology
CN107767879A (en) Audio conversion method and device based on tone color
Achterkamp et al. The influence of vicarious experience provided through mobile technology on self-efficacy when learning new tasks
MX2020002093A (en) Genealogical entity resolution system and method.
JP2009259085A (en) Biometric authentication system and authentication method
Abbasi et al. Statistical, spectral and graph representations for video-based facial expression recognition in children
EA202092400A1 (en) METHOD FOR OBTAINING VOICE-DEPENDENT SMALL-SIZED HIGH-LEVEL ACOUSTIC SPEECH FEATURES
Akasaka et al. Model-free template reconstruction attack with feature converter
RU2005102541A (en) METHOD FOR HUMAN IDENTIFICATION BY ITS BIOMETRIC IMAGE
Poh et al. A biometric menagerie index for characterising template/model-specific variation
GB2594873A (en) Biometric processes, apparatus and machine-readable mediums
MX2021000543A (en) System and method for genealogical entity resolution.
FR3076378B1 (en) METHOD OF FORMING A NEURON NETWORK FOR THE RECOGNITION OF A SEQUENCE OF CHARACTERS AND ASSOCIATED METHOD OF RECOGNITION
EP3226166A3 (en) Systems and methods for authentication based on human teeth pattern
Dyakonova About reguLAtion of the comPetent PArticiPAntS in ProcedurAL LegiSLAtion
RU2008142215A (en) METHOD FOR FORMING ELECTRONIC BIOMETRIC IDENTIFICATION
WO2019193362A3 (en) Determining a clinical outcome for a subject suffering from a macular degenerative disease
CN113707294B (en) Psychological evaluation method based on dynamic video data
Ting et al. Performance analysis of single and combined bit-planes feature extraction for recognition in face expression database
Gulhane et al. A review on surgically altered face images recognition using multimodal bio-metric features
Noriega Encoding auditory-visual interactions in a neural model of sensory abnormalities in autism
Stubbs The universal declaration of human rights at 70
EA201800121A3 (en) METHOD OF IDENTIFICATION OF A HUMAN APPROACH ON EEG RESPONSE TO MIXED IMAGES