Face recognition using total loss function on face database with ID photos

• Published in: Optics and laser technology Vol. 110; pp. 227 - 233
• Main Authors: Cui, Dongshun, Zhang, Guanghao, Hu, Kai, Han, Wei, Huang, Guang-Bin
• Format: Journal Article
• Language: English
• Published: Kidlington Elsevier Ltd 01.02.2019; Elsevier BV
• Subjects: Algorithms; Artificial neural networks; Face recognition; Face recognition benchmark; Facial recognition technology; Feature extraction; Human performance; ID photos; Image processing systems; Pattern recognition systems; Pictures; Real-life face recognition system; Total loss function; ID photos; Face recognition; Total loss function; Real-life face recognition system; Face recognition benchmark
• ISSN: 0030-3992; 1879-2545
• DOI: 10.1016/j.optlastec.2017.10.016

•We collected a new face database (FDID) with ID photos for each subject.•We compared our new face database with the existing face databases.•We proposed a specified and novel face recognition method for our database.•The proposed algorithm was evaluated on our database and face video sequences. With the development of deep neural networks, researchers have developed lots of algorithms related to face and achieved comparable results to human-level performance on several databases. However, few feature extraction models work well in the real world when the subject which is to be recognized has limited samples, for example, only one ID photo can be obtained before the face recognition task. To our best knowledge, there is no face database which contains ID photos and pictures from the real world for a subject simultaneously. To fill this gap, we collected 100 celebrities’ ID photos and their about 1000 stills or life pictures and formed a face database called FDID. Besides, we proposed a novel face recognition algorithm and evaluated it with this new database on the real-life videos.