Skip to content

Dataset · zhifeixie

Audio dataset for ASR in complex acoustic conditions

zhifeixie/Voices-in-the-Wild-2M

Large-scale English and Chinese audio corpus for training robust ASR systems.

What's inside

Contains English and Chinese audio data for the automatic-speech-recognition (ASR) task. The dataset covers 7 classical acoustic phenomena (noise, far-field, echo, distortion, dropouts, etc.) and 54 complex combined scenarios.

What you can do with it

  • Training ASR models robust to background noise and acoustic interference
  • Testing and evaluating speech recognition quality in real-world conditions
  • Developing voice assistants for complex acoustic environments
Tasks
Automatic speech recognition
Languages
English, Chinese
Downloads this month
1.7K
Likes
52

Models trained on it

How to load it

from datasets import load_dataset

ds = load_dataset("zhifeixie/Voices-in-the-Wild-2M")