neoneye/simon-arc-solve-boundingbox-v19

该数据集的核心任务是识别边界框（bounding box）。随着版本的更新，图像大小、边界框的类型（如填充、空心等）以及任务的复杂性逐渐增加。最新版本（Version 19）将RLE压缩响应替换为原始像素响应，并设置了最大图像大小为14。

Simons ARC (abstraction & reasoning corpus) solve boundingbox version 19 is an image processing dataset designed to train models to identify and generate various types of bounding boxes. The dataset includes multiple versions, each with different image sizes and bounding box types, including filled, hollow, inner, and outer bounding boxes, as well as background transformations. The latest version 19 replaces RLE compressed responses with raw pixel responses, with a maximum image size of 14.