dataset-jaw-movements
收藏资源简介:
该数据集包含奶牛在放牧时咀嚼动作的音频记录和标签。通过无线麦克风记录了三头荷斯坦奶牛在食用短和高苜蓿以及短和高雀麦时的声音。两位放牧行为专家识别并标记了每次咀嚼动作的开始、结束和类型:咬合、咀嚼和咀嚼-咬合(复合动作)。对于每个原始音频段,计算了四个知名特征:振幅、持续时间、零交叉和包络对称。这些特征可用于构建自动分类咀嚼动作的方法。
This dataset comprises audio recordings and labels of chewing actions by cows during grazing. The sounds of three Holstein cows consuming short and tall alfalfa, as well as short and tall brome grass, were recorded using wireless microphones. Two grazing behavior experts identified and labeled the start, end, and type of each chewing action: bite, chew, and chew-bite (composite action). For each original audio segment, four well-known features were calculated: amplitude, duration, zero-crossing, and envelope symmetry. These features can be utilized to construct methods for automatically classifying chewing actions.
数据集概述
基本信息
- 标题: Audio recordings dataset of grazing jaw movements in dairy cattle
- 作者: Sebastián R. Vanrell, José O. Chelotti, Leandro A. Bugnon, H. Leonardo Rufiner, Diego H. Milone, Emilio A. Laca, Julio R. Galli
- 发表期刊: Data in Brief
- DOI: https://doi.org/10.1016/j.dib.2020.105623
数据集描述
- 内容: 包含52个WAV格式的音频文件,记录了三头荷斯坦奶牛在不同草地上(短和高的苜蓿,短和高的牛毛草)的咀嚼动作声音。
- 标签: 52个TXT文件,由两位草食行为专家标注,记录了每次咀嚼动作的开始、结束和类型(咬合、咀嚼、咬合-咀嚼复合动作)。
- 特征计算: 对每个音频片段计算了四个特征:振幅、持续时间、零交叉和包络对称性。
- 应用: 可用于构建自动分类咀嚼动作的方法,监测和分析奶牛的草食行为。
数据集内容
- 音频文件: 52个WAV文件
- 标签文件: 52个TXT文件
- 总结文件: 1个CSV文件




