Papers › Vibravox: A Dataset of French Speech Captured with Body-conduction Audio Sensors
Vibravox: A Dataset of French Speech Captured with Body-conduction Audio Sensors
16 Jul 2024arXiv:2407.11828archive 2025-07-28
Julien Hauret, Malo Olivier, Thomas Joubaud, Christophe Langrenne, Sarah Poirée, Véronique Zimpfer, Éric Bavu
Vibravox is a dataset compliant with the General Data Protection Regulation (GDPR) containing audio recordings using five different body-conduction audio sensors: two in-ear microphones, two bone conduction vibration pickups, and a laryngophone. The dataset also includes audio data from an airborne microphone used as a reference. The Vibravox corpus contains 45 hours per sensor of speech samples and physiological sounds recorded by 188 participants under different acoustic conditions imposed by a high order ambisonics 3D spatializer. Annotations about the recording conditions and linguistic transcriptions are also included in the corpus. We conducted a series of experiments on various speech-related tasks, including speech recognition, speech enhancement, and speaker verification. These experiments were carried out using state-of-the-art models to evaluate and compare their performances on signals captured by the different audio sensors offered by the Vibravox dataset, with the aim of gaining a better grasp of their individual characteristics.
Code
Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.
Code Syntology ran Syntology
Not run by Syntology. Nothing on this page verifies that the listed code works.
Tasks
Datasets
Introduced by this paper, per the archive.
Results from the paper archive 2025-07-28
| Task | Dataset | Model | Metric | Value | Rank at snapshot | Leaderboard | Report |
| Automatic Phoneme Recognition |
VibraVox (forehead accelerometer) |
medium wav2vec2.0 |
Test PER |
0.046 |
#1 of 1 |
Archive leaderboard |
report |
| Automatic Phoneme Recognition |
VibraVox (headset microphone) |
medium wav2vec2.0 |
Test PER |
0.028 |
#1 of 1 |
Archive leaderboard |
report |
| Automatic Phoneme Recognition |
VibraVox (rigid in-ear microphone) |
medium wav2vec2.0 |
Test PER |
0.045 |
#1 of 1 |
Archive leaderboard |
report |
| Automatic Phoneme Recognition |
VibraVox (soft in-ear microphone) |
medium wav2vec2.0 |
Test PER |
0.041 |
#1 of 1 |
Archive leaderboard |
report |
| Automatic Phoneme Recognition |
VibraVox (temple vibration pickup) |
medium wav2vec2.0 |
Test PER |
0.142 |
#1 of 1 |
Archive leaderboard |
report |
| Automatic Phoneme Recognition |
VibraVox (throat microphone) |
medium wav2vec2.0 |
Test PER |
0.073 |
#1 of 1 |
Archive leaderboard |
report |
| Bandwidth Extension |
VibraVox (forehead accelerometer) |
Configurable EBEN (M=4, P=4, Q=4) |
EER (ECAPA2) |
0.0183 |
#1 of 1 |
Archive leaderboard |
report |
| Bandwidth Extension |
VibraVox (forehead accelerometer) |
Configurable EBEN (M=4, P=4, Q=4) |
Noresqua-MOS |
4.250 |
#1 of 1 |
Archive leaderboard |
report |
| Bandwidth Extension |
VibraVox (forehead accelerometer) |
Configurable EBEN (M=4, P=4, Q=4) |
PER (wav2vec2) |
0.091 |
#1 of 1 |
Archive leaderboard |
report |
| Bandwidth Extension |
VibraVox (forehead accelerometer) |
Configurable EBEN (M=4, P=4, Q=4) |
STOI |
0.855 |
#1 of 1 |
Archive leaderboard |
report |
| Bandwidth Extension |
VibraVox (rigid in-ear microphone) |
Configurable EBEN (M=4, P=2, Q=4) |
EER (ECAPA2) |
0.0364 |
#1 of 1 |
Archive leaderboard |
report |
| Bandwidth Extension |
VibraVox (rigid in-ear microphone) |
Configurable EBEN (M=4, P=2, Q=4) |
Noresqua-MOS |
4.285 |
#1 of 1 |
Archive leaderboard |
report |
| Bandwidth Extension |
VibraVox (rigid in-ear microphone) |
Configurable EBEN (M=4, P=2, Q=4) |
PER (wav2vec2) |
0.084 |
#1 of 1 |
Archive leaderboard |
report |
| Bandwidth Extension |
VibraVox (rigid in-ear microphone) |
Configurable EBEN (M=4, P=2, Q=4) |
STOI |
0.877 |
#1 of 1 |
Archive leaderboard |
report |
| Bandwidth Extension |
VibraVox (soft in-ear microphone) |
Configurable EBEN (M=4, P=2, Q=4) |
EER (ECAPA2) |
0.0488 |
#1 of 1 |
Archive leaderboard |
report |
| Bandwidth Extension |
VibraVox (soft in-ear microphone) |
Configurable EBEN (M=4, P=2, Q=4) |
Noresqua-MOS |
4.331 |
#1 of 1 |
Archive leaderboard |
report |
| Bandwidth Extension |
VibraVox (soft in-ear microphone) |
Configurable EBEN (M=4, P=2, Q=4) |
PER (wav2vec2) |
0.087 |
#1 of 1 |
Archive leaderboard |
report |
| Bandwidth Extension |
VibraVox (soft in-ear microphone) |
Configurable EBEN (M=4, P=2, Q=4) |
STOI |
0.868 |
#1 of 1 |
Archive leaderboard |
report |
| Bandwidth Extension |
VibraVox (temple vibration pickup) |
Configurable EBEN (M=4, P=1, Q=4) |
EER (ECAPA2) |
0.1622 |
#1 of 1 |
Archive leaderboard |
report |
| Bandwidth Extension |
VibraVox (temple vibration pickup) |
Configurable EBEN (M=4, P=1, Q=4) |
Noresqua-MOS |
3.632 |
#1 of 1 |
Archive leaderboard |
report |
| Bandwidth Extension |
VibraVox (temple vibration pickup) |
Configurable EBEN (M=4, P=1, Q=4) |
PER (wav2vec2) |
0.391 |
#1 of 1 |
Archive leaderboard |
report |
| Bandwidth Extension |
VibraVox (temple vibration pickup) |
Configurable EBEN (M=4, P=1, Q=4) |
STOI |
0.763 |
#1 of 1 |
Archive leaderboard |
report |
| Bandwidth Extension |
VibraVox (throat microphone) |
Configurable EBEN (M=4, P=2, Q=4) |
EER (ECAPA2) |
0.0847 |
#1 of 1 |
Archive leaderboard |
report |
| Bandwidth Extension |
VibraVox (throat microphone) |
Configurable EBEN (M=4, P=2, Q=4) |
Noresqua-MOS |
3.862 |
#1 of 1 |
Archive leaderboard |
report |
| Bandwidth Extension |
VibraVox (throat microphone) |
Configurable EBEN (M=4, P=2, Q=4) |
PER (wav2vec2) |
0.179 |
#1 of 1 |
Archive leaderboard |
report |
| Bandwidth Extension |
VibraVox (throat microphone) |
Configurable EBEN (M=4, P=2, Q=4) |
STOI |
0.834 |
#1 of 1 |
Archive leaderboard |
report |
| Speaker Verification |
VibraVox (forehead accelerometer) |
ECAPA2 |
Test EER |
0.009 |
#1 of 1 |
Archive leaderboard |
report |
| Speaker Verification |
VibraVox (forehead accelerometer) |
ECAPA2 |
Test min-DCF |
0.06 |
#1 of 1 |
Archive leaderboard |
report |
| Speaker Verification |
VibraVox (headset microphone) |
ECAPA2 |
Test EER |
0.0026 |
#1 of 1 |
Archive leaderboard |
report |
| Speaker Verification |
VibraVox (headset microphone) |
ECAPA2 |
Test min-DCF |
0.02 |
#1 of 1 |
Archive leaderboard |
report |
| Speaker Verification |
VibraVox (rigid in-ear microphone) |
ECAPA2 |
Test EER |
0.0316 |
#1 of 1 |
Archive leaderboard |
report |
| Speaker Verification |
VibraVox (rigid in-ear microphone) |
ECAPA2 |
Test min-DCF |
0.21 |
#1 of 1 |
Archive leaderboard |
report |
| Speaker Verification |
VibraVox (soft in-ear microphone) |
ECAPA2 |
Test EER |
0.0172 |
#1 of 1 |
Archive leaderboard |
report |
| Speaker Verification |
VibraVox (soft in-ear microphone) |
ECAPA2 |
Test min-DCF |
0.10 |
#1 of 1 |
Archive leaderboard |
report |
| Speaker Verification |
VibraVox (temple vibration pickup) |
ECAPA2 |
Test EER |
0.08 |
#1 of 1 |
Archive leaderboard |
report |
| Speaker Verification |
VibraVox (temple vibration pickup) |
ECAPA2 |
Test min-DCF |
0.58 |
#1 of 1 |
Archive leaderboard |
report |
| Speaker Verification |
VibraVox (throat microphone) |
ECAPA2 |
Test EER |
0.0353 |
#1 of 1 |
Archive leaderboard |
report |
| Speaker Verification |
VibraVox (throat microphone) |
ECAPA2 |
Test min-DCF |
0.20 |
#1 of 1 |
Archive leaderboard |
report |
Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.
Methods
Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections