ianisdev/imagenet_augmented
收藏Hugging Face2025-07-24 更新2025-10-25 收录
下载链接：
https://hf-mirror.com/datasets/ianisdev/imagenet_augmented
下载链接
链接失效反馈官方服务：
资源简介：
这是一个增强版的ImageNet子集，用于图像分类任务，包含了多种增强方法，如传统增强、Mixup、MIA Mix、AutoAugment、LSB-level bit noise、BigGAN class-conditional samples、VQ-VAE reconstructions和fusion等。数据集旨在研究不同增强方法对大规模图像分类模型泛化能力的影响。
This dataset is an augmented version of a subset of ImageNet, used for image classification tasks, containing various augmentation methods such as traditional augmentation, Mixup, MIA Mix, AutoAugment, LSB-level bit noise, BigGAN class-conditional samples, VQ-VAE reconstructions, and fusion. The dataset is designed to study the impact of different augmentation methods on the generalization ability of large-scale image classification models.
提供机构：
ianisdev
