Bee Audio Object Detection

The dataset was created with a focus on detecting bees using sound as the primary input. The audio recordings were collected from the entrances of 25 beehives located in three separate apiaries situated in San Jose, Cupertino, and Gilroy, California, USA.

The audio recordings were made uniform by adjusting their length to 10 seconds. Each recording was closely reviewed twice and categorized as having bee sounds or not having bee sounds based on the presence or absence of such sounds. The audio clips that contain bee sounds may have various background noises, such as birds, airplanes, cars, and other sounds.

该数据集的构建旨在专注于通过声音作为主要输入来检测蜜蜂。音频记录采集自位于美国加利福尼亚州圣何塞、库比蒂诺和吉尔罗伊的三个独立蜂场中25个蜂箱的入口。通过对音频长度进行调整，确保所有记录均统一为10秒。每段录音均经过两次仔细审查，并根据是否存在蜜蜂声音将其分类为包含蜜蜂声音或不包含蜜蜂声音。包含蜜蜂声音的音频片段可能伴有各种背景噪音，例如鸟鸣、飞机、汽车以及其他声响。