Abstract
Despite the importance of few-shot learning, the lack of labeled training data in the real world makes it extremely challenging for existing machine learning methods because this limited dataset does not well represent the data variance. In this research, we suggest employing a generative approach using variational autoencoders (VAEs), which can be used specifically to optimize few-shot learning tasks by generating new samples with more intra-class variations on the Labeled Faces in the Wild (LFW) dataset. The purpose of our research is to increase the size of the training dataset using various methods to improve the accuracy and robustness of the few-shot face recognition. Specifically, we employ the VAE generator to increase the size of the training dataset, including the basic and the novel sets while utilizing transfer learning as the backend. Based on extensive experimental research, we analyze various data augmentation methods to observe how each method affects the accuracy of face recognition. The face generation method based on VAEs with perceptual loss can effectively improve the recognition accuracy rate to 96.47% using both the base and the novel sets.
Highlights
The explosion of big data has provided enough training samples in the real world to facilitate the development of deep learning performance [1,2,3]
The generative model variational autoencoders (VAEs) is trained by the CASIA-WebFace dataset [99], which has a total of 10,575 people and a total of 494,414 images
This is because ordinary VAE tries to minimize the pixel-by-pixel loss between two images, whereas pixel-based loss does not contain perceptual and spatially related information
Summary
The explosion of big data has provided enough training samples in the real world to facilitate the development of deep learning performance [1,2,3]. As a result of the development of high-performance computing devices such as graphics processing units (GPUs) and CPU clusters in recent years, the training of large-scale deep learning models has been greatly improved for big data feature learning. Deep learning models are often successful with millions of model parameters and a large amount of labeled big data available for training. Many applications of this kind of deep learning in face recognition can only be realized on the premise of having a large amount of labeled data. In real life, due to restrictions such as data security management and labor costs, it is impractical to obtain such a large amount of labeled data
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
