Face recognition model training method and device

Embodiments of the present application provide a face recognition model training method and device, wherein the method includes: acquiring image data; using the improved feature extraction model to extract image feature information from image data, wherein the improved feature extraction model is an...

Full description

Saved in:
Bibliographic Details
Main Author LUO PEIYUAN
Format Patent
LanguageChinese
English
Published 11.01.2019
Subjects
CALCULATING
COMPUTING
COUNTING
HANDLING RECORD CARRIERS
PHYSICS
PRESENTATION OF DATA
RECOGNITION OF DATA
RECORD CARRIERS
Online AccessGet full text

Cover

More Information
Summary:Embodiments of the present application provide a face recognition model training method and device, wherein the method includes: acquiring image data; using the improved feature extraction model to extract image feature information from image data, wherein the improved feature extraction model is an improved mobilenet V2 model; normalizing the image feature information by the improved normalization layer, and obtaining the normalized image feature information; obtaining the feature vectors d from the normalized image feature information by the improved embedding layer; training a human face recognition model according to the feature vector. Because the improved mobilenet V2 model is used as feature extraction model and the normalized layer and embedded layer are modified accordingly, the model structure is simplified and the processing precision is taken into account. Then the improved model is used to train the face recognition model, which solves the technical problem that the existing methods are inefficie
Bibliography:Application Number: CN201810746086