Shad0wKillar/pizza_steak_sushi
收藏资源简介:
该数据集是从原始Food101数据集中随机抽取20%的自定义子集,专门筛选出三类食物:披萨、牛排和寿司。它设计用于轻量级图像分类实验和PyTorch教程。数据集已预先分为训练集和测试集目录,划分比例与原始Food101数据集相同(默认75/25的训练/测试比例),但缩小为20%的随机抽样。总共有600张图像,其中训练集450张(披萨154张,寿司150张,牛排146张),测试集150张(披萨46张,寿司46张,牛排58张)。数据集采用标准的PyTorch ImageFolder兼容布局组织。
This dataset is a custom, 20% random subset of the original Food101 dataset, specifically filtered to contain only three classes: Pizza, Steak, and Sushi. It is designed for lightweight image classification experiments and PyTorch tutorials. The data is pre-split into `train` and `test` directories. The split ratios mirror the original Food101 dataset splits (which defaults to a 75/25 train/test ratio), but scaled down to a 20% random sampling. Total Images: 600. Train Split: 450 images (Pizza: 154 images, Sushi: 150 images, Steak: 146 images). Test Split: 150 images (Pizza: 46 images, Sushi: 46 images, Steak: 58 images). The dataset is organized in a standard PyTorch `ImageFolder` compatible layout.




