voa-engines/streaming-vad-dataset

该数据集包含经过语音活动检测（VAD）处理的葡萄牙语（pt-br）音频片段。每个音频片段包括一个.wav格式的音频文件和相关元数据，如自动生成的原始转录、持续时间、开始和结束时间。此数据集可用于手动审查和纠正转录、评估VAD参数以及针对葡萄牙语的自动语音识别（ASR）模型进行微调。

This dataset contains Portuguese (pt-br) audio segments processed with Voice Activity Detection (VAD). Each audio segment includes a .wav file and associated metadata such as automated raw transcription, duration, start and end times. The dataset is intended for manual review and correction of transcriptions, evaluation of VAD parameters, and fine-tuning of Portuguese ASR models.