Bibliographic record
Abstract
Millions of different people talk to each other, and no two people sound exactly the same. Yet, whomever we are listening to, we expect to easily understand what he or she has to say. Somehow, we adjust to each new talker’s voice and hear the “same” speech sounds. Until recently, differences between voices were viewed as a perceptual problem interfering with speech perception. Recent developments, however, have shown that familiar voices can facilitate speech. Speech-perception models can no longer dismiss talkers’ voices merely as carriers for speech, and models currently struggle to understand the relation between vocal identity and the content of speech. The present thesis contributed to this discussion by examining familiar talkers, whose identities have been encoded into listeners’ memories. Chapter 2 studied familiar faces’ and voices’ contribution to audiovisual speech processing, and found that different listeners may focus more strongly on learning either a familiar talker’s face or voice, but will recall what they have learned in response to that talker’s voice, not face. Chapter 3 examined self-speech, and discovered that we do receive a familiar-talker speech-processing advantage from hearing our own recorded voice, but only so far as we can identify self-voice; when voices are obscured by noise, we receive an equivalent advantage for all voices of our own sex. Chapter 4 confirmed a relation between speech familiarity and accurate talker identification. Taken together, the data presented in this thesis support a model of speech perception in which listeners encode talkers’ identities inclusive of both idiosyncratic speech production and vocal qualities and, when processing speech, recall as many of a talker’s identifying characteristics as can be usefully applied to an incoming speech signal. These findings contribute to our understanding of how we utilize talker identity in perceiving speech.
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 | 0.003 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.001 | 0.002 |
| Scholarly communication | 0.002 | 0.003 |
| Open science | 0.000 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.004 | 0.000 |
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".