TomasGuija/LatinFontsSVGs

SVG字体数据集是一个精心策划的矢量字体字形数据集，以SVG文件形式存储，专为生成建模、结构化矢量图形和字体合成研究而设计。该数据集用于论文DesigNet: Learning to Draw Vector Graphics as Designers Do的开发与评估。主要特征包括：SVG字形表示；仅使用三次贝塞尔曲线，无二次贝塞尔曲线；字形坐标归一化到[0,1]范围；已预处理可直接用于模型管道；提供训练、验证和测试分割；分割在字体家族级别进行，每个字体家族仅出现在一个分割中；数据经过多次人工审核以确保质量，但可能仍存在一些缺陷、异常或不完美的字体。

The SVG Font Dataset is a curated dataset of vector font glyphs stored as SVG files, designed for research in generative modeling, structured vector graphics, and typography synthesis. It was created for the development and evaluation of the paper DesigNet: Learning to Draw Vector Graphics as Designers Do. Key features include: SVG glyph representations; no quadratic Bezier curves are present, only cubic curves; glyph coordinates are normalized to the [0,1] range; preprocessing has been applied for direct use in the model pipeline; train, validation, and test splits are provided; splits are performed at the family level, with each font family appearing entirely in only one split; data has been manually reviewed multiple times for quality control, although some faulty, unusual, or imperfect fonts may still remain.