Automatic forced alignment on child speech: Directions for improvement
Bibliographic record
Abstract
Phonetic analysis is labor intensive, limiting the amount of data that can be considered. Recently, automated techniques (e.g., forced alignment based on Automatic Speech Recognition - ASR) have emerged allowing for much larger-scale analyses. For adult speech, forced alignment can be accurate even when the phonetic transcription is automatically generated, allowing for large-scale phonetic studies. However, such analyses remain difficult for children's speech, where ASR methods perform more poorly. The present study used a trainable forced aligner that performs well on adult speech to examine the effect of four factors on alignment accuracy of child speech: (1) Corpus - elicited speech (multiple children) versus spontaneous speech (single child); (2) Pronunciation dictionary - standard adult versus customized; (3) Training data - adult lab speech, corpus-specific child speech, all child speech, or a combination of child and adult speech; (4) Segment type - voiceless stops, voiceless sibilants, and vowels. Automatic and manual segmentations were compared. Greater accuracy was observed with (1) elicited speech, (2) customized pronunciations, (3) training on child speech, and (4) stops. These factors increase the utility of analyzing children's speech production using forced alignment, potentially allowing researchers to ask questions that otherwise would require weeks or months of manual-segmentation.
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 | 0.021 | 0.045 |
| Meta-epidemiology (narrow) | 0.002 | 0.001 |
| Meta-epidemiology (broad) | 0.002 | 0.001 |
| Bibliometrics | 0.002 | 0.002 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.003 | 0.006 |
| Open science | 0.004 | 0.001 |
| Research integrity | 0.003 | 0.002 |
| Insufficient payload (model declined to judge) | 0.021 | 0.007 |
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".