SYNAuG: Exploiting Synthetic Data for Data Imbalance Problems

arXiv:2308.00994v3 Announce Type: replace-cross
Abstract: Data imbalance in training data often leads to biased predictions from trained models, which in turn causes ethical and social issues. A straightforward solution is to carefully curate training data, but given the enormous scale of modern neural networks, this is prohibitively labor-intensive and thus impractical. Inspired by recent developments in generative models, this paper explores the potential of synthetic data to address the data imbalance problem. To be specific, our method, dubbed SYNAuG, leverages …

abstract arxiv cs.cv cs.lg data ethical labor leads modern networks neural networks predictions scale social social issues solution synthetic synthetic data training training data type