A survey of synthetic data augmentation methods in computer vision

arXiv:2403.10075v1 Announce Type: cross
Abstract: The standard approach to tackling computer vision problems is to train deep convolutional neural network (CNN) models using large-scale image datasets which are representative of the target task. However, in many scenarios, it is often challenging to obtain sufficient image data for the target task. Data augmentation is a way to mitigate this challenge. A common practice is to explicitly transform existing images in desired ways so as to create the required volume and variability …

abstract arxiv augmentation cnn computer computer vision convolutional neural network cs.cv cs.gr cs.lg data datasets however image image data image datasets network neural network scale standard survey synthetic synthetic data train type vision