Oxford 102 Flower

该数据集名为Flowers102，专为对花卉种类进行细粒度分类而设计，包含了102个不同的类别。在仅有的每类20张示例图片的情况下，分类器成功学习了如何区分这102种花卉。所执行的任务是细粒度图像分类。

This dataset, designated as Flowers102, is purpose-built for fine-grained classification of flower species, encompassing 102 distinct categories. With merely 20 sample images per category, the classifier successfully learned to differentiate among these 102 flower species. The task conducted herein is fine-grained image classification.

搜集汇总

背景与挑战

背景概述

Oxford 102 Flower数据集包含102种英国常见花卉的图片，每类40-258张不等，具有丰富的尺度、姿态和光照变化。数据集还提供图像分割、χ2距离等评估数据，适用于花卉分类研究。