Speech-processing researcher at WPI working on full-duplex speech-to-speech models and tool-calling voice agents
2
papers here
2
reviewed by Evolve
0
first-author
0
corresponding
—
works · OpenAlex
6
h-index · OpenAlex
Bibliometric figures come from OpenAlex (2026-10-15) and cover the author's entire public record, not just the papers listed here. Reproducibility signal comes from Evolve's own reviews.
Research profile
Output—
Not yet scored: complete output aggregates are unavailable in this snapshot. This axis would measure publication volume and productivity trends over time; the only signals on file are yearly publication counts for 2018–2026 and an h-index of 6.
Impact—
Not yet scored: citation-based impact aggregates are insufficient in this snapshot. This axis would measure citation counts, citation velocity, and influence on follow-up work; beyond an i10-index of 5, no reliable impact metrics are available yet.
Focus0/100
The stored focus score is 0, but the raw field distribution suggests concentration rather than scatter: fieldDist splits evenly (0.5/0.5) between two near-identical labels — 'speech processing and conversational AI' and 'speech processing / spoken dialogue systems (cs.CL)' — and topic counts are speech-dominated (Speech Recognition and Synthesis: 22; Speech and Audio Processing: 21). Treat the 0 as a likely labeling artifact.
Collaboration83/100
Collaboration scores 83, the strongest axis, supported by 64 distinct co-authors. Both catalog papers are team efforts with middle authorship and no corresponding role, consistent with work inside a stable speech-engineering group alongside recurring co-authors such as Edresson Casanova, Zhehuai Chen, and Jagadeesh Balam.
Activity61/100
Activity scores 61. countsByYear shows steady but modest output: 2 works in 2018, a peak of 6 in 2020, then 2–4 per year through 2024, with recent3yWorks at 7 (2023: 3, 2024: 2, 2025: 2, 2026: 3 to date). A sustained, moderate cadence rather than high-volume publishing.
Reproducibility38/100
Reproducibility scores 38, the mean across the 2 papers Evolve reviewed. Readiness is rated medium for the VoiceChat full-duplex model and low for the frontend-backend tool-call architecture, which was flagged for absent significance testing and low reproducibility; released artifacts and evaluation rigor are the main gaps.
Research summary
Viet Anh Trinh is a speech-processing researcher at Worcester Polytechnic Institute whose work centers on full-duplex speech-to-speech models and tool-calling voice agents. This profile is name-resolved from public bibliographic data (OpenAlex; ORCID 0000-0002-1660-6627). The snapshot records an h-index of 6 and an i10-index of 5, with a publishing record spanning 2018–2026: 2 works in 2018, a peak of 6 in 2020, and 7 works over the most recent three years. Topic assignment is heavily concentrated in speech: Speech Recognition and Synthesis (22), Speech and Audio Processing (21), speech and dialogue systems (7), with smaller clusters in phonetics and phonology (4) and music and audio processing (4); the field split is even (0.5/0.5) between speech processing and conversational AI and spoken dialogue systems (cs.CL). Two 2026 arXiv papers are in the Evolve catalog: NemotronLabs VoiceChat, an open full-duplex speech-to-speech model with native tool calling and strong turn-taking, and a frontend-backend architecture that adds tool calls to duplex speech models via delegation tokens. Both are team publications with middle authorship and no corresponding role, situated in a network of 64 distinct co-authors (collaboration score 83). Reproducibility averages 38 across the two reviewed papers.
Collaboration pattern
Collaboration is the strongest axis (83), with 64 distinct co-authors recorded. Both Evolve-reviewed papers are large-team arXiv efforts in which Viet Anh Trinh holds a middle authorship position without corresponding role. Recurring co-authors include Edresson Casanova, Zhehuai Chen, Jagadeesh Balam, Elena Rastorgueva, Nikhil Srihari, Travis Bartley, Slyne Deng, Oluwatobi Olabiyi, Sanjay Chauhan, and Chen Chen, indicating sustained work within a stable speech-engineering group.
Research keywords
Highlights
Open full-duplex speech-to-speech model with native tool calling
Token-based delegation architecture for tool calls in duplex speech models
Timeline
- 2018Earliest works in the snapshot appear (2 publications), starting a publishing record that runs through 2026.
- 2020Most active year on record with 6 publications, within a topic profile dominated by speech recognition and synthesis and speech and audio processing.
- 20224 works published, continuing a steady presence in speech processing research.
- 2024Two works each in 2024 and 2025 — a quieter stretch before renewed output in 2026.
- 20263 works to date, including two Evolve-reviewed arXiv papers on full-duplex speech-to-speech models with tool calling.
Papers in the Evolve catalog
Frequent co-authors
About this profile
This profile is AI-structured by Evolve.Pub from public bibliographic data (OpenAlex) and Evolve's own paper reviews. It is not written or endorsed by the researcher unless the profile is claimed. Radar dimensions are computed from real bibliometrics with fixed formulas — raw figures are shown alongside; missing dimensions mean insufficient data, never zero. Errors? Claim the profile to correct or hide it.
- Bibliometrics come from an OpenAlex snapshot (h-index 6, i10-index 5, yearly counts); citation coverage may lag recent work.
- Output and impact axes are unscored because the snapshot's underlying aggregates were insufficient.
- Evolve has reviewed only 2 of this researcher's papers; verdicts and the reproducibility score cover just those works.
- Identity is name-resolved from public bibliographic data (OpenAlex, ORCID 0000-0002-1660-6627); confirm before citing.