Fotopersbureau De Boer Training Set on Scene Detection

This dataset was created as part of the projects HisVis and <em>Fotografisch Geheugen </em>conducted by the Noord-Hollands Archief and University of Amsterdam in the Netherlands. These projects examined to what extent Computer Vision, and more specifically, scene detection could be applied to a collection of historical press photographs. A computer was trained to automatically recognize specific scenes on historical press photographs, like a ‘protest’, ‘marriage’, ‘shopping street’ or ‘baseball game’. The key aspect of scene recognition is to identify the place in which the objects seat. The specific aim of the enrichments provided by scene detection was to benefit users of the archive and cultural historians studying historical photographs. The training set contains historical press photographs of the collection of Fotopersbureau De Boer (1945-2005). This file includes the training data used for the training of a scene detection model as well as a model to detect whether a picture was taken indoors or outdoors. The model cards, data sheet, and label sheet include more information on the dataset. - examples.tar.gz contains example images for each label. The label sheet contains more information - indoor_out.tar.gz contains the training set for the indoor / outdoor model - scene_detection.tar.gz contains the training set for the scene detection model. - no_description_found.tar.gz contains images that were not linked to any class during annotation. - HisVis2-0.1.beta.tar.gz contains the source code for the software. It stems from this GitHub repo: https://github.com/melvinwevers/HisVis2 - models.tar.gz contains the trained models. These should be placed in the repo. The models are not included here because of their size. More information on the data is available in the data sheet: <br>本数据集由荷兰北荷兰档案馆（Noord-Hollands Archief）与阿姆斯特丹大学（University of Amsterdam）共同实施的HisVis项目及*Fotografisch Geheugen*（摄影记忆）项目所创建。上述两项研究聚焦于探究计算机视觉（Computer Vision）技术——尤其是场景检测（scene detection）技术——可在多大程度上应用于历史新闻摄影藏品库。 研究团队开发了可自动识别历史新闻照片中特定场景的计算机模型，可识别的场景涵盖“抗议活动”“婚礼现场”“商业街”与“棒球比赛”等。场景识别的核心在于明确物体所处的空间位置。本次通过场景检测技术实现的标注增强工作，核心目标是为档案馆用户与研究历史照片的文化史学者提供赋能。 本次训练集（training set）采用了Fotopersbureau De Boer（德波尔摄影事务所，1945-2005）馆藏的历史新闻照片。本文件包含了用于训练场景检测模型，以及用于训练图像室内/室外场景分类模型的全部训练数据。模型卡片（model cards）、数据集说明文档（data sheet）与标签表（label sheet）中包含了本数据集的更多详细信息。 各压缩包的具体内容如下： - examples.tar.gz：包含每个标签对应的示例图像，更多细节可参见标签表（label sheet） - indoor_out.tar.gz：包含室内/室外场景分类模型的训练数据集（training set） - scene_detection.tar.gz：包含场景检测模型的训练数据集（training set） - no_description_found.tar.gz：包含标注阶段未关联至任何类别的图像 - HisVis2-0.1.beta.tar.gz：包含配套软件的源代码，其源自下述GitHub仓库：https://github.com/melvinwevers/HisVis2 - models.tar.gz：包含训练完成的模型文件。因模型文件体积较大，未随本数据包一同上传，用户需将解压后的模型文件放置于对应仓库目录中。 更多数据集相关信息可查阅数据集说明文档（data sheet）。