AudioEchoes
收藏AudioEchoes 数据集
数据集描述
AudioEchoes 数据集包含多种回声环境的音频文件。每个文件都由先进的语音识别系统转录,旨在帮助开发和评估语音转文本算法,特别是在识别和处理语音信号中的回声方面。数据集包括多种场景,如大厅中的回声、大房间中的回声和户外环境中的回声,标签指示存在的回声类型。
CSV 内容预览
"Filename","Transcription","Labels" "audio_echo_hall.wav","The meeting will commence at nine in the morning.","hall_echo" "audio_echo_room.wav","Please pass the salt and pepper to the guests.","large_room_echo" "audio_echo_outdoor.wav","The birds are singing loudly today, arent they lovely?","outdoor_echo" "audio_echo_hall.wav","We need to order more supplies for the event next week.","hall_echo" "audio_echo_room.wav","Lets discuss the quarterly sales figures over lunch.","large_room_echo"
数据来源
该数据集使用 Infinite Dataset Hub 和 microsoft/Phi-3-mini-4k-instruct 生成,查询为 speech to text:
- 数据集生成页面: https://huggingface.co/spaces/infinite-dataset-hub/infinite-dataset-hub?q=speech+to+text&dataset=AudioEchoes&tags=transcription,+speech+recognition,+deep+learning
- 模型: https://huggingface.co/microsoft/Phi-3-mini-4k-instruct
- 更多数据集: https://huggingface.co/datasets?other=infinite-dataset-hub




