metadata
dataset_info:
features:
- name: image
dtype: image
- name: speaker
dtype: string
- name: speakerID
dtype: int64
- name: label
dtype: int64
- name: dataset
dtype: string
- name: valence
dtype: float64
- name: arousal
dtype: float64
- name: domination
dtype: float64
- name: arousal_norm
dtype: float64
- name: valence_norm
dtype: float64
- name: file_paths
dtype: string
splits:
- name: train
num_bytes: 189294383.46336207
num_examples: 2340
- name: validation
num_bytes: 53285482.648731746
num_examples: 687
- name: test
num_bytes: 55879477.02408564
num_examples: 572
download_size: 343068424
dataset_size: 298459343.13617945
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*