USDA White Mountain National Forest Volume 1 (2014 - 2024)

This volume's release consists of 325099 media files captured by autonomous wildlife monitoring devices under the project, USDA White Mountain National Forest. The attached files listed below include several CSV files that provide information about the data release. The file, "media.csv" provides the metadata about the media, such as filename and date/time of capture. The actual media files are housed within folders under the volume's "child items" as compressed files. A critical CSV file is "dictionary.csv", which describes each CSV file, including field names, data types, descriptions, and the relationship of each field to fields in other CSV files. Some of the media files may have been "tagged" or "annotated" by either humans or by machine learning models, identifying wildlife targets within the media. If so, this information is stored in "annotations.csv" and "modeloutputs.csv", respectively. To protect privacy, all personally identifiable information (PII) have been removed, locations have been "blurred" by bounding boxes, and media featuring sensitive taxa or humans have been omitted. To enhance data reuse, the sbRehydrate() function in the AMMonitor R package will download files and re-create the original AMMonitor project (database + media files). See source code at https://code.usgs.gov/vtcfwru/ammonitor.

本次发布卷包含由美国农业部（United States Department of Agriculture, USDA）怀特山国家森林项目下的自主野生动物监测设备采集的325099份媒体文件。本次发布附带的文件包含若干CSV格式文件，用于说明本次数据发布的相关信息。其中"media.csv"文件包含媒体文件的元数据，例如文件名及采集日期时间。实际的媒体文件以压缩包形式存储于本次发布卷的"子项"下属文件夹中。核心CSV文件为"dictionary.csv"，该文件对所有CSV文件进行说明，涵盖字段名称、数据类型、字段含义，以及各字段与其他CSV文件字段的关联关系。部分媒体文件已由人工或机器学习模型进行"标记"或"注释"，以识别媒体中的野生动物目标；若存在此类标注信息，则分别存储于"annotations.csv"与"modeloutputs.csv"文件中。为保护隐私，本次发布已移除所有个人可识别信息（Personally Identifiable Information, PII），通过边界框对位置信息进行"模糊化"处理，并删除了包含敏感类群或人类的媒体文件。为提升数据复用性，AMMonitor R软件包中的sbRehydrate()函数可下载相关文件并重建原始AMMonitor项目（数据库+媒体文件）。源代码可访问以下链接获取：https://code.usgs.gov/vtcfwru/ammonitor。