CASE数据集

CASE数据集是由ADAPT SFI研究中心创建的大型户外场景数据集，包含10000张来自Cityscapes数据集的街景图像，每张图像均附有手动标注的广告植入候选空间。数据集的创建过程涉及雇佣志愿者，使用VGG Image Annotator工具在图像中标记可能的广告植入位置，确保广告与场景视角匹配。CASE数据集主要应用于视频编辑领域，旨在帮助编辑在视频帧中系统地选择合适的广告植入空间，以增强视频内容的个性化和吸引力。

The CASE Dataset is a large-scale outdoor scene dataset developed by the ADAPT SFI Research Centre. It contains 10,000 street view images sourced from the Cityscapes dataset, with each image paired with manually annotated candidate spaces for advertising placement. The dataset's creation process involved recruiting volunteers to use the VGG Image Annotator tool to mark potential advertising placement positions in the images, ensuring that the advertising placement aligns with the perspective of the corresponding scene. Primarily applied in the field of video editing, the CASE dataset aims to help editors systematically select appropriate advertising placement spaces within video frames to enhance the personalization and attractiveness of video content.