A data-centric approach to class-specific bias in image data augmentation

arXiv:2403.04120v1 Announce Type: new
Abstract: Data augmentation (DA) enhances model generalization in computer vision but may introduce biases, impacting class accuracy unevenly. Our study extends this inquiry, examining DA's class-specific bias across various datasets, including those distinct from ImageNet, through random cropping. We evaluated this phenomenon with ResNet50, EfficientNetV2S, and SWIN ViT, discovering that while residual models showed similar bias effects, Vision Transformers exhibited greater robustness or altered dynamics. This suggests a nuanced approach to model selection, emphasizing bias mitigation. …

abstract accuracy arxiv augmentation bias biases class computer computer vision cs.cv data data-centric datasets image image data imagenet model generalization random resnet50 study through type vision