JawLink: An In-Ear Optical Interface for Orofacial Movement Monitoring and Speech Detection
Bibliographic record
Abstract
Voice prostheses such as the tracheoesophageal prosthesis (TEP) and electrolarynx (EL) restore speech after total laryngectomy but rely on manual activation, limiting autonomy and usability. This study investigates an in-ear proximity sensing approach for hands-free detection of speech intent, leveraging soft tissue deformation within the external auditory canal during mandibular motion. Ten healthy participants were fitted with bilateral, custom-fitted ear devices incorporating infrared sensors to record deformation patterns during structured speech and non-speech tasks. Signals were preprocessed using low-pass filtering, wavelet transforms, and temporal smoothing. A feature set combining time- and frequency-domain descriptors was extracted and ranked using minimum redundancy-maximum relevance selection criteria. Classifiers, including boosted trees and wide neural networks, were trained to distinguish speech from confounding orofacial behaviors. Results revealed consistent pre-phonatory and post-phonatory activation phases in the proximity signal, indicating detection of articulatory intent beyond acoustic onset. The best model for detecting speech with respect to baseline achieved an average F1 score of 84.04% $\pm ~$ 4.48% in a leave-one-subject-out cross-validation. An additional validation demonstrated a high F1 score of 94.49% $\pm ~$ 2.49% for distinguishing speech from clustered non-speech activities (e.g. chewing, yawning, smiling). These results support the robustness of the proposed system against false triggers. This work provides the first evidence that in-ear deformation signals can be used for robust speech detection in healthy controls, setting the stage for real-time, hands-free control of voice prostheses. The system offers a discreet, wearable platform for assistive communication technologies in patients with profound voice impairments.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.001 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.001 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.000 |
| Insufficient payload (model declined to judge) | 0.008 | 0.002 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".