Datasets:

Modalities:
Audio
Text
Formats:
parquet
AVE-Dataset / README.md
AdnanElAssadi's picture
Upload dataset
f6eb93b verified
metadata
dataset_info:
  features:
    - name: video
      dtype: video
    - name: audio
      dtype:
        audio:
          sampling_rate: 16000
    - name: label
      dtype:
        class_label:
          names:
            '0': Accordion
            '1': Acoustic guitar
            '2': Baby cry, infant cry
            '3': Banjo
            '4': Bark
            '5': Bus
            '6': Cat
            '7': Chainsaw
            '8': Church bell
            '9': Clock
            '10': Female speech, woman speaking
            '11': Fixed-wing aircraft, airplane
            '12': Flute
            '13': Frying (food)
            '14': Goat
            '15': Helicopter
            '16': Horse
            '17': Male speech, man speaking
            '18': Mandolin
            '19': Motorcycle
            '20': Race car, auto racing
            '21': Rodents, rats, mice
            '22': Shofar
            '23': Toilet flush
            '24': Train horn
            '25': Truck
            '26': Ukulele
            '27': Violin, fiddle
    - name: youtube_id
      dtype: string
    - name: start_seconds
      dtype: int32
  splits:
    - name: train
      num_bytes: 5150175797
      num_examples: 3312
    - name: test
      num_bytes: 664329126
      num_examples: 402
  download_size: 6359281087
  dataset_size: 5814504923
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
      - split: test
        path: data/test-*