Mitsua/vrm-color-concept-550k

VRM Color Concept 550K是一个用于训练动漫风格文本到图像或多模态模型的数据集，所有使用的资产均为CC0或适当许可，确保无版权问题。数据集包括图像、元数据和渲染过程，所有内容均遵循CC BY-NC 4.0许可证，允许非商业用途的使用和改编，但需给予适当署名。数据集通过自动化的渲染系统生成，包括环境HDRI图像、姿势、表情、相机角度和后处理效果的自动变化，所有标题均基于预定义的规则生成，避免了使用预训练标题生成器可能带来的版权问题。

This is a dataset for training anime-style text-to-image or any text and image multimodal models without copyright/licensing concerns. All assets/materials are CC0 or properly licensed, and no pretrained models or any AI models are used to build this dataset. Images, metadata, and the dataset itself are licensed under CC BY-NC 4.0 for non-commercial purposes. The dataset uses VRM models, HDRI images, pose data, and a customized renderer. All rendering and captioning are done through an automated system to ensure no knowledge leakage of copyrighted works.