human_help_nursing_home
收藏资源简介:
该数据集是一个多模态数据集,主要包含音频数据及相关联的数值特征。数据集由75个训练样本构成,总大小约为598MB。每个样本包含以下字段:audio字段存储音频数据;au_secours和a_laide字段均为浮点数列表,可能表示与音频相关的某种数值特征或标注(这两个字段名称在法语中均意为救命或帮助,暗示数据集可能与紧急呼救检测或特定语音指令识别任务相关);speaker_id字段为整数类型,用于标识不同的说话人。数据集适用于音频处理、语音指令识别、多说话人语音分析或紧急场景语音检测等任务。
This dataset is a multimodal dataset primarily consisting of audio data and associated numerical features. It comprises 75 training samples with a total size of approximately 598MB. Each sample includes the following fields: the audio field stores the audio data; the au_secours and a_laide fields are both lists of floating-point numbers, likely representing some numerical features or annotations related to the audio (both field names mean help or rescue in French, suggesting the dataset may be related to emergency call detection or specific voice command recognition tasks); the speaker_id field is an integer type used to identify different speakers. The dataset is suitable for tasks such as audio processing, voice command recognition, multi-speaker speech analysis, or emergency scenario voice detection.
根据您提供的数据集详情页面信息,以下是对该数据集的总结:
数据集概述
数据集名称:human_help_nursing_home
数据集来源:Hugging Face Datasets(KandoCare 组织)
主要特征
该数据集包含以下三个特征字段:
- audio:音频数据(数据类型为
audio) - au_secours:浮点数列表(
list类型,元素为float64) - a_laide:浮点数列表(
list类型,元素为float64) - speaker_id:说话人标识(数据类型为
int64)
数据划分
数据集仅包含一个划分:
- 训练集(train):
- 样本数量:75 个
- 数据大小:约 598.66 MB(字节数:598,655,318)
数据文件
- 配置名称:
default - 训练集数据文件路径:
data/train-*
其他信息
- 下载大小:约 598.56 MB(字节数:598,557,276)
- 数据集总大小:约 598.66 MB(字节数:598,655,318)




