Small facial image dataset augmentation using conditional GANs based on incomplete edge feature input.

ABSTRACT: Image data collection and labelling is costly or difficult in many real applications. Generating diverse and controllable images using conditional generative adversarial networks (GANs) for data augmentation from a small dataset is promising but challenging as deep convolutional neural networks need a large training dataset to achieve reasonable performance in general. However, unlabeled and incomplete features (e.g., unintegral edges, simplified lines, hand-drawn sketches, discontinuous geometry shapes, etc.) can be conveniently obtained through pre-processing the training images and can be used for image data augmentation. This paper proposes a conditional GAN framework for facial image augmentation using a very small training dataset and incomplete or modified edge features as co

SUBMITTER: Hung SK 

PROVIDER: S-EPMC8627232 | biostudies-literature | 2021

REPOSITORIES: biostudies-literature