cemag/tl-caxton

TL-Caxton 3D打印质量评估数据集是一个专注于增材制造/3D打印领域的图像数据集。它包含了4048张RGB图像以及与之对应的数值注释。数据集支持多种任务类型，包括流量回归、挤出质量分类、视觉语言建模和视觉问答。数据集分为训练集、验证集和测试集，分别占总样本的84.2%、8.2%和7.6%。此外，数据集还提供了用于生成自然语言描述的JSON模板文件，以支持视觉语言模型的训练。

The TL-Caxton 3D Printing Quality Assessment Dataset is an image dataset focused on the field of additive manufacturing/3D printing. It contains 4,048 RGB images along with corresponding numerical annotations. The dataset supports multiple task types, including flow rate regression, extrusion quality classification, vision-language modeling, and visual question answering. The dataset is split into training, validation, and test sets, accounting for 84.2%, 8.2%, and 7.6% of the total samples, respectively. Additionally, the dataset provides JSON template files for generating natural language descriptions to support the training of vision-language models.