A survey of automated data augmentation for image classification: Learning to compose, mix, and generate

TH Cheung, DY Yeung - IEEE transactions on neural networks …, 2023 - ieeexplore.ieee.org
Data augmentation is an effective way to improve the generalization of deep learning
models. However, the underlying augmentation methods mainly rely on handcrafted …

A Survey of Automated Data Augmentation for Image Classification: Learning to Compose, Mix, and Generate.

TH Cheung, DY Yeung - IEEE Transactions on Neural Networks and …, 2023 - europepmc.org
Data augmentation is an effective way to improve the generalization of deep learning
models. However, the underlying augmentation methods mainly rely on handcrafted …

A Survey of Automated Data Augmentation for Image Classification: Learning to Compose, Mix, and Generate

TH Cheung, DY Yeung - IEEE transactions on neural networks … - pubmed.ncbi.nlm.nih.gov
Data augmentation is an effective way to improve the generalization of deep learning
models. However, the underlying augmentation methods mainly rely on handcrafted …