{"success":true,"data":{"author":"Julien Hauret, Malo Olivier, Thomas Joubaud, Christophe Langrenne, Sarah Poiree, Veronique Zimpfer, Eric Bavu (Cnam LMSSC, French-German Research Institute of Saint-Louis)","bucketAccessType":"public","bucketName":"Cnam-LMSSC/vibravox","bucketPrefix":"","category":"Speech & Audio","cloudProvider":"huggingface","createdAt":"2026-11-14T14:15:18.054Z","description":"Vibravox is a general-purpose audio dataset of French speech captured simultaneously with body-conduction transducers and conventional microphones: a headset microphone near the mouth as reference, a forehead vibration sensor, a foam-embedded in-ear microphone, a rigid in-ear earpiece microphone, a temple vibration pickup and a laryngophone. Body-conducted speech is robust to ambient noise but band-limited and distorted, so the dataset is used for automatic speech recognition (speech-to-text and speech-to-phoneme), audio bandwidth extension, speech enhancement, speaker verification and identification, and voice cloning from non-conventional sensors. It has four subsets: speech_clean, speech_noisy (speech recorded in noise), speechless_clean and speechless_noisy (sensor responses without speech, for noise and physiological-sound studies). This repository holds them as four folders of Parquet shards, for example 255 files in speech_clean (measured 2026-11-14; the Hub reports about 564 GB). Load with datasets.load_dataset('Cnam-LMSSC/vibravox', 'speech_clean', streaming=True); a pip package 'vibravox' provides training code. Released under CC BY 4.0; cite Hauret et al., Speech Communication 2025.","displayTitle":"Vibravox: French speech recorded with body-conduction sensors","doi":"10.57967/hf/2727","id":"629b5463-e616-5ba6-9158-6bc83e2d4c32","license":"CC-BY-4.0","licenseEvidenceUrl":"https://huggingface.co/datasets/Cnam-LMSSC/vibravox","objectCountEstimate":360,"probeStatus":"ok","referenceUrl":"https://arxiv.org/abs/2407.11828","region":"us","schemaCard":{"accessHints":[{"language":"python-datasets","snippet":"from datasets import load_dataset\nds = load_dataset(\"Cnam-LMSSC/vibravox\")"},{"language":"python-hub","snippet":"from huggingface_hub import snapshot_download\nsnapshot_download(repo_id=\"Cnam-LMSSC/vibravox\", repo_type=\"dataset\")"}],"caveats":["HuggingFace metadata: author=Cnam-LMSSC, license:cc-by-4.0, size_categories:10K<n<100K, format:parquet, downloads=12322"],"format":"mixed","generatedAt":"2026-11-14T14:15:23.408Z","layout":{"example":"speech_clean/test-00000-of-00030.parquet","pattern":"{repoId}/{filename}"},"mountPath":"/mnt/archive","sampleHeaders":[{"bytes":"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","encoding":"hex","key":"speech_clean/test-00000-of-00030.parquet","truncatedAt":512}],"sampleObjects":["speech_clean/test-00000-of-00030.parquet","speech_clean/test-00001-of-00030.parquet","speech_clean/test-00002-of-00030.parquet","speech_clean/test-00003-of-00030.parquet","speech_clean/test-00004-of-00030.parquet"],"sidecars":[{"bytes":"---\nannotations_creators:\n- expert-generated\nlanguage_creators:\n- crowdsourced\n- expert-generated\nlanguage:\n- fr\nlicense: cc-by-4.0\nmultilinguality:\n- monolingual\nsize_categories:\n- 100K<n<1M\nsource_datasets: []\ntask_categories:\n- audio-to-audio\n- automatic-speech-recognition\n- audio-classification\n- text-to-speech\ntask_ids:\n- speaker-identification\npretty_name: Vibravox\nviewer: true\ndataset_info:\n- config_name: speech_clean\n  features:\n  - name: audio.headset_microphone\n    dtype: audio\n  - name: audio.forehead_accelerometer\n    dtype: audio\n  - name: audio.soft_in_ear_microphone\n    dtype: audio\n  - name: audio.rigid_in_ear_microphone\n    dtype: audio\n  - name: audio.temple_vibration_pickup\n    dtype: audio\n  - name: audio.throat_microphone\n    dtype: audio\n  - name: gender\n    dtype: string\n  - name: speaker_id\n    dtype: string\n  - name: sentence_id\n    dtype: int64\n  - name: duration\n    dtype: float64\n  - name: raw_text\n    dtype: string\n  - name: normalized_text\n    dtype: string\n  - name: phonemized_text\n    dtype: string\n  splits:\n  - name: train\n    num_bytes: 100144385419.375\n    num_examples: 20981\n  - name: validation\n    num_bytes: 11821970622.625\n    num_examples: 2523\n  - name: test\n    num_bytes: 14647423280.0\n    num_examples: 3064\n  download_size: 124418585390\n  dataset_size: 126613779322.0\n- config_name: speech_noisy\n  features:\n  - name: audio.headset_microphone\n    dtype: audio\n  - name: audio.forehead_accelerometer\n    dtype: audio\n  - name: audio.soft_in_ear_microphone\n    dtype: audio\n  - name: audio.rigid_in_ear_microphone\n    dtype: audio\n  - name: audio.temple_vibration_pickup\n    dtype: audio\n  - name: audio.throat_microphone\n    dtype: audio\n  - name: gender\n    dtype: string\n  - name: speaker_id\n    dtype: string\n  - name: sentence_id\n    dtype: int64\n  - name: duration\n    dtype: float64\n  - name: raw_text\n    dtype: string\n  - name: normalized_text\n    dtype: string\n  - name: phonemized_text\n    dtype: string\n  splits:\n  - name: train\n    num_bytes: 5978781164.5\n    num_examples: 1220\n  - name: validation\n    num_bytes: 647300251.0\n    num_examples: 132\n  - name: test\n    num_bytes: 859092188.0\n    num_examples: 175\n  download_size: 7471066223\n  dataset_size: 7485173603.5\n- config_name: speechless_clean\n  features:\n  - name: audio.headset_microphone\n    dtype: audio\n  - name: audio.forehead_accelerometer\n    dtype: audio\n  - name: audio.soft_in_ear_microphone\n    dtype: audio\n  - name: audio.rigid_in_ear_microphone\n    dtype: audio\n  - name: audio.temple_vibration_pickup\n    dtype: audio\n  - name: audio.throat_microphone\n    dtype: audio\n  - name: gender\n    dtype: string\n  - name: speaker_id\n    dtype: string\n  - name: duration\n    dtype: float64\n  splits:\n  - name: train\n    num_bytes: 8512005740.0\n    num_examples: 149\n  - name: validation\n    num_bytes: 1028286672.0\n    num_examples: 18\n  - name: test\n    num_bytes: 1199717890.0\n    num_examples: 21\n  download_size: 9548480336\n  dataset_size: 10740010302.0\n- config_name: speechless_noisy\n  features:\n  - name: audio.headset_microphone\n    dtype: audio\n  - name: audio.forehead_accelerometer\n    dtype: audio\n  - name: audio.soft_in_ear_microphone\n    dtype: audio\n  - name: audio.rigid_in_ear_microphone\n    dtype: audio\n  - name: audio.temple_vibration_pickup\n    dtype: audio\n  - name: audio.throat_microphone\n    dtype: audio\n  - name: gender\n    dtype: string\n  - name: speaker_id\n    dtype: string\n  - name: duration\n    dtype: float64\n  splits:\n  - name: train\n    num_bytes: 24723250192.0\n    num_examples: 149\n  - name: validation\n    num_bytes: 2986606278.0\n    num_examples: 18\n  - name: test\n    num_bytes: 3484522468.0\n    num_examples: 21\n  download_size: 30881658818\n  dataset_size: 31194378938.0\nconfigs:\n- config_name: speech_clean\n  data_files:\n  - split: train\n    path: speech_clean/train-*\n  - split: validation\n    path: speech_clean/validation-*\n  - split: test\n    path: speech_clean/test-*\n- config_name: speech_noisy\n  data_files:\n  - split: train\n    path: speech_noisy/train-*\n  - split: validation\n    path: speech_noisy/valida","encoding":"utf8","key":"README.md","name":"README.md","size":35608,"truncated":true}],"sourceObjectKey":"speech_clean/test-00000-of-00030.parquet","starterCell":"from datasets import load_dataset\nds = load_dataset(\"Cnam-LMSSC/vibravox\")\nprint(ds)","starterSamplePath":"speech_clean/test-00000-of-00030.parquet","totalObjects":360},"summary":"Vibravox is a general-purpose audio dataset of French speech captured simultaneously with body-conduction transducers and conventional microphones: a headset microphone near the…","title":"Vibravox: French speech recorded with body-conduction sensors","updatedAt":"2026-11-14T17:41:44.509Z","visibility":"public","websiteUrl":"https://vibravox.cnam.fr"}}