metadata
dataset_info:
features:
- name: audio
sequence: int16
- name: label
dtype: string
splits:
- name: train
num_bytes: 66033056
num_examples: 727
download_size: 57573927
dataset_size: 66033056
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
license: mit
task_categories:
- feature-extraction
- zero-shot-classification
The Nonverbal Vocalization Dataset, also known as the Vocal Characterizer, is a human nonverbal vocal sound dataset that was crowdsourced from the South Korean public and consists of 56.7 hours of brief clips from 1419 speakers. The dataset also contains metadata about sex, age, noise level, and speech quality. "Teeth-chattering," "teeth-grinding," "tongue-clicking," "coughing," "yawning," "throat clearing," "sighing," "lip-popping," "lip-smacking," "panting," "crying," "laughing," "sneezing," "moaning," and "screaming" are among the 16 classifications of included human nonverbal sound data.
gadget