Nexdata/German_Children_Spontaneous_Speech_Data
收藏Hugging Face2024-01-26 更新2024-03-04 收录
下载链接：
https://hf-mirror.com/datasets/Nexdata/German_Children_Spontaneous_Speech_Data
下载链接
链接失效反馈官方服务：
资源简介：
---
task_categories:
- automatic-speech-recognition
language:
- de
---
# Dataset Card for Nexdata/German_Children_Spontaneous_Speech_Data
## Description
The 97 Hours - German Child's Spontaneous Speech Data, manually screened and processed. Annotation contains transcription text, speaker identification, gender and other informantion. This dataset can be applied in speech recognition (acoustic model or language model training), caption generation, voice content moderation and other AI algorithm research.
For more details, please refer to the link: https://www.nexdata.ai/datasets/1299?source=Huggingface
# Specifications
## Format
16k Hz, 16 bit, wav, mono channel;
## Age
12 years old and younger children;
## Content category
including self-media, conversation, live, lecture, variety show;
## Language
German;
## Annotation
annotation for the transcription text, speaker identification, gender;
## Accuracy
Word Accuracy Rate (WAR) at least 98%.
# Licensing Information
Commercial License
提供机构：
Nexdata
原始信息汇总
数据集卡片 Nexdata/German_Children_Spontaneous_Speech_Data
描述
97小时 - 德国儿童自发语音数据，经过人工筛选和处理。标注包含转录文本、说话者识别、性别等信息。该数据集可应用于语音识别（声学模型或语言模型训练）、字幕生成、语音内容审核等AI算法研究。
规格
格式
16k Hz，16位，wav格式，单声道；
年龄
12岁及以下儿童；
内容类别
包括自媒体、对话、直播、讲座、综艺节目；
语言
德语；
标注
转录文本、说话者识别、性别标注；
准确性
单词准确率（WAR）至少98%。
许可信息
商业许可
