metadata

annotations_creators:
  - expert-generated
language_creators:
  - crowdsourced
  - expert-generated
language:
  - fr
license: cc-by-4.0
multilinguality:
  - monolingual
task_categories:
  - audio-to-audio
  - automatic-speech-recognition
  - audio-classification
  - text-to-speech
task_ids:
  - speaker-identification
pretty_name: Vibravox
dataset_info:
  config_name: speech_clean
  features:
    - name: audio.headset_microphone
      dtype: audio
    - name: audio.forehead_accelerometer
      dtype: audio
    - name: audio.soft_in_ear_microphone
      dtype: audio
    - name: audio.rigid_in_ear_microphone
      dtype: audio
    - name: audio.temple_vibration_pickup
      dtype: audio
    - name: audio.throat_microphone
      dtype: audio
    - name: gender
      dtype: string
    - name: speaker_id
      dtype: string
    - name: sentence_id
      dtype: int64
    - name: duration
      dtype: float64
    - name: raw_text
      dtype: string
    - name: normalized_text
      dtype: string
    - name: phonemized_text
      dtype: string
  splits:
    - name: test
      num_bytes: 4879701252
      num_examples: 3064
  download_size: 4541794656
  dataset_size: 4879701252
configs:
  - config_name: speech_clean
    data_files:
      - split: test
        path: speech_clean/test-*

Dataset Card

Description

This dataset features a speech-enhanced version of the test split from the speech_clean subset of the Vibravox Dataset. It is not intended for training.

Enhancement procedure

The Bandwidth extension task has been individually achieved for each sensor using configurable EBEN (arXiv link) models available at https://huggingface.co/Cnam-LMSSC/vibravox_EBEN_models.

Ressources

Results for speech-to-phoneme and speaker verification tasks using this enhanced dataset are available in our paper, and code is available at the jhauret/vibravox Github repository.