jaddai/openbrush-van-gogh

OpenBrush Van Gogh数据集是从OpenBrush-75K数据集中精选出的子集，专门收录文森特·梵高的所有作品，共包含1,889张图像。每张图像都配有结构化的视觉语言模型（VLM）标注，这些标注基于Qwen3-VL-30B-A3B模型生成，涵盖多个方面：标签、主题、动作、场景、情绪、风格、光照、颜色和构图。数据集旨在支持图像到文本、文本到图像和图像分类等任务，特别适用于艺术分析和生成模型训练。梵高的作品涵盖了他的现实主义时期（如荷兰风景和农民场景）和后印象主义突破期（如阿尔勒、圣雷米和奥维尔的作品），标注详细描述了他的厚涂笔触、调色板变化和主题处理。数据集基于MIT许可证，艺术品本身属于公共领域（所有艺术家在1954年前去世）。

OpenBrush Van Gogh is a curated subset of the OpenBrush-75K dataset, containing all Vincent van Gogh works with structured VLM captions. It includes 1,889 images, each annotated with detailed captions generated by the Qwen3-VL-30B-A3B model. The captions follow a structured schema with fields such as TAGS, SUBJECT, ACTION, SETTING, MOOD, STYLE, LIGHTING, COLOR, and COMPOSITION. The dataset is designed for tasks like image-to-text, text-to-image, and image-classification, particularly useful for art analysis and generative model training. Van Goghs catalog spans from his Realism period (Dutch landscapes, peasant scenes) to Post-Impressionist breakthroughs (Arles, Saint-Rémy, Auvers). The dataset is licensed under MIT, with underlying artworks in the public domain (all artists died before 1954).