iMaterialist Fashion Attribute

iFashion-Attribute数据库中的所有图像均由Wish提供，1,012,947，9,897和39,706图像分别分为训练，验证和测试集。它有228个细粒度的时尚属性级别的课程，形成了8个高级时装团体从时尚行业专业定义。每个图像都有多个标签。每个组都有许多时尚类别，数量范围从3 (“性别” 组) 到105 (“类别” 组)。

All images in the iFashion-Attribute database are provided by Wish. A total of 1,012,947, 9,897, and 39,706 images are partitioned into the training, validation, and test sets, respectively. It comprises 228 fine-grained fashion attribute-level categories, which form 8 high-level fashion groups as defined by fashion industry professionals. Each image is annotated with multiple labels. Each group contains multiple fashion categories, with the count ranging from 3 (for the "Gender" group) to 105 (for the "Category" group).

搜集汇总

背景与挑战

背景概述

iMaterialist Fashion Attribute是一个大规模的时尚属性数据集，包含超过100万张图像，分为训练、验证和测试集，用于细粒度时尚属性分类任务。数据集涵盖228个细粒度属性类别，组织成8个高级时装组，每个图像具有多标签标注，适用于图像分类和时尚分析研究。该数据集由康奈尔大学、码隆科技和Google AI Research于2019年发布，旨在支持时尚行业的计算机视觉应用。