DeepFashion2 服饰数据集

DeepFashion2 是一个包括衣物检测、姿态估计、分割及检索 4 个任务的通用基准。该数据集包括 80.1 万个服装类目 (clothes item)，每个类目都有丰富的标注，如 style, scale, viewpoint, occlusion, bounding box, dense landmark, mask 。此外它还包含 87.3 万个 commercial- consumer 服装对。

DeepFashion2 is a universal benchmark encompassing four core tasks: clothing detection, pose estimation, segmentation, and retrieval. The dataset includes 801,000 clothing items, each paired with rich annotations such as style, scale, viewpoint, occlusion, bounding box, dense landmark, and mask. Additionally, it contains 873,000 commercial-consumer clothing pairs.

搜集汇总

背景与挑战

背景概述

DeepFashion2是一个包含80.1万个服装类目和87.3万个商业-消费者服装对的服饰数据集，提供丰富的标注信息如风格、比例、视角等，支持衣物检测、姿态估计、分割及检索4个任务。该数据集适用于图像处理、识别和语义分割领域的研究与应用。