WHAM!48kHz noise dataset

This is a mirror of the WHAM!48kHz noise dataset. The original files were segmented and converted from WAV to Opus to reduce the size and accelerate streaming.

• Sampling rate: 48 kHz
• Channels: 2
• Format: Opus
• Splits:
  • Train: 59 hours, 21216 segments, files 000 to 188
  • Validation: 12 hours, 4444 segments, files 189 to 225
  • Test: 7 hours, 2613 segments, files 226 to 249
• License: CC BY-NC 4.0
• Source: http://wham.whisper.ai/
• Paper: WHAM!: Extending Speech Separation to Noisy Environments

Usage

import io

import soundfile as sf
from datasets import Features, Value, load_dataset

for item in load_dataset(
    "philgzl/wham",
    split="train",
    streaming=True,
    features=Features({"audio": Value("binary"), "name": Value("string")}),
):
    print(item["name"])
    buffer = io.BytesIO(item["audio"])
    x, fs = sf.read(buffer)
    # do stuff...

Citation

@inproceedings{wichern2019wham,
  title = {{WHAM!}: {Extending} speech separation to noisy environments},
  author = {Wichern, Gordon and Antognini, Joe and Flynn, Michael and Zhu, Licheng Richard and McQuinn, Emmett and Crow, Dwight and Manilow, Ethan and Roux, Jonathan Le},
  booktitle = {Proc. Interspeech},
  pages = {1368--1372},
  year = {2019},
}