USC speech and audio AI researcher: non-verbal vocalization ASR and encoder selection for large audio-language models
2
papers here
2
reviewed by Evolve
0
first-author
0
corresponding
—
works · OpenAlex
4
h-index · OpenAlex
Bibliometric figures come from OpenAlex (2026-10-09) 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—
Output data is not yet available for this profile. This axis would measure total publication volume, venue quality, and production trends, so no score can be reported.
Impact—
Impact data is insufficient for scoring. This axis would measure citation counts, h-index trajectory, and influence on downstream work, none of which can be estimated here.
Focus0/100
The radar focus score is 0, the lowest band: the field distribution splits evenly (0.5/0.5) between machine learning for audio and speech processing and speech processing / computational paralinguistics, and OpenAlex topics span five areas — speech recognition and synthesis (24), speech and audio processing (18), music and audio processing (5), multimodal machine learning (4), and emotion recognition (3). In practice the spread stays inside the audio-speech domain.
Collaboration56/100
Collaboration scores 56. The OpenAlex record lists 16 distinct coauthors against 25 works in the last three years, and 12 recurring collaborators including Hung-yi Lee, Hsin-Min Wang, and Hung-Shin Lee. Both reviewed 2026 papers carry middle-author positions without corresponding authorship, consistent with a contributor role inside team-produced speech systems rather than solo or lead-corresponding work.
Activity95/100
Activity scores 95, the profile's strongest axis. Yearly works climb steadily from 2 in 2023 to 5 in 2024, 8 in 2025, and 12 in 2026 — a six-fold increase over four years — with 25 works in the past three years. The trajectory indicates a rapidly accelerating publication cadence centered on 2025-2026.
Reproducibility55/100
Reproducibility scores 55, averaged over the 2 papers Evolve has reviewed; both are rated medium reproducibility readiness. The NVVSpeech system paper is flagged for an unexplained local-versus-official score gap and thin statistics, while CUES reports strong results but is validated on a single small backbone and one benchmark — leaving replication scope untested.
Research summary
Wenze Ren is a researcher at the University of Southern California working where machine learning meets speech and audio processing. This is an early-stage profile resolved from public OpenAlex data: the record shows an h-index of 4 and an i10-index of 1, 25 works over the past three years, and a sharply rising annual output of 2 works in 2023, 5 in 2024, 8 in 2025, and 12 in 2026. Topic activity concentrates on speech recognition and synthesis (24 records) and speech and audio processing (18), with smaller footprints in music and audio processing (5), multimodal machine learning applications (4), and emotion and mood recognition (3); the field split is even, 50/50 between machine learning for audio and speech processing and speech processing / computational paralinguistics. Evolve has reviewed two 2026 arXiv papers listing Ren as a middle author: an NVVSpeech Challenge Track 1 system that reaches rank 4 with a score of 63.86 using sqrt-plus-uniform category sampling for non-verbal vocalization ASR, and CUES, a correlation-guided method for choosing encoders in multi-encoder large audio-language models. Both carry medium reproducibility readiness, and the review mean is 55 across the two papers; verdicts praise honesty while flagging narrow validation.
Collaboration pattern
With 16 distinct coauthors across 25 works in three years, Ren works in team settings: the frequent-collaborator list includes Hung-yi Lee, Hsin-Min Wang, Hung-Shin Lee, Kuo-Hsuan Hung, and Pei-Jun Liao, alongside peers such as Shangyue Jia and Jingru Ma. Both reviewed 2026 papers place Ren in a middle authorship position without corresponding authorship, indicating a contributor role within larger speech research groups rather than an independent lab-lead pattern.
Research keywords
Highlights
Track 1 rank 4 in the NVVSpeech Challenge
CUES: correlation-guided encoder selection for audio-language models
Output accelerating six-fold since 2023
OpenAlex-recorded works rise from 2 in 2023 to 5 in 2024, 8 in 2025, and 12 in 2026, with 25 works over the past three years and an h-index of 4 at this early stage.
Timeline
- 2023First visible OpenAlex record: 2 works published, the starting point of the publication trail.
- 2024Output grows to 5 works, more than doubling the previous year.
- 20258 works recorded, continuing the steep upward trend in publication activity.
- 2026Most productive year yet with 12 works, including both Evolve-reviewed arXiv papers — the NVVSpeech Challenge Track 1 system (rank 4, 63.86) and the CUES encoder-selection method.
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.
- Bibliometric figures (h-index 4, i10-index 1, works by year, coauthor counts) come from OpenAlex only; the output and impact dimensions lack sufficient data and are not scored.
- Evolve has reviewed only two papers, both 2026 arXiv preprints with middle authorship; verdicts cover these alone.
- Identity is resolved from public bibliographic data; 'Wenze Ren' may correspond to more than one researcher.
- The reproducibility mean (55) rests on just 2 reviewed papers, both rated medium readiness, and may shift as more work is assessed.