Early-career researcher at Université de Yaoundé I applying HGNN cross-modal transfer to low-resource Yemba speech
2
papers here
2
reviewed by Evolve
2
first-author
0
corresponding
—
works · OpenAlex
0
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 is not scored yet: the OpenAlex snapshot lists only two 2026 arXiv preprints, with h-index and i10-index at 0. This axis would measure total publication volume across venues and years once more records accumulate.
Impact—
Impact is not scored yet: no citation counts are available for the two 2026 preprints, and h-index stands at 0. This axis would track citations, uptake by other speech-technology work, and downstream reuse of the HGNN methods.
Focus0/100
The focus score is 0/100 in this snapshot. The field distribution splits evenly, 0.5 to speech processing / multimodal representation learning and 0.5 to speech processing / computational linguistics, across only 2 works; OpenAlex topics likewise cluster around speech recognition and synthesis, speech and audio processing, and graph neural networks.
Collaboration22/100
The collaboration score is 22/100, based on a small OpenAlex network: 2 distinct coauthors recorded, with Marie Tahon and Paulin Melatagia Yonta as the recurring partners across both 2026 first-author preprints. This points to an early-stage, concentrated bilateral working pattern rather than an extensive multi-institution collaboration network.
Activity32/100
The activity score is 32/100: OpenAlex counts 2 works in the last three years, both appearing in 2026, and the counts-by-year record contains only that single 2026 entry. Output is thus recent and concentrated rather than sustained; the current pace reflects two arXiv preprints with no longer publication history to indicate continuity.
Reproducibility55/100
The reproducibility score is 55/100, computed over 2 papers with a mean readiness of 55. Evolve rates both HGNN preprints as medium: methods and the Yemba low-resource setting are described, but reviews flag a missing acoustic-model baseline, single-run evaluation, and absent frugal baselines as gaps a reproduction would need to close.
Research summary
Yannick Yomie Nzeuhang is an early-stage researcher at Université de Yaoundé I in Cameroon, working at the intersection of speech processing and graph-based machine learning with a focus on low-resource African languages. The Evolve catalog holds two 2026 first-author arXiv preprints, both built on heterogeneous graph neural networks (HGNN) to improve speech representation learning for Yemba. One applies explicit text-to-audio cross-modal knowledge transfer to isolated word recognition; the other frames word-level speech-text alignment as frugal HGNN link prediction. Evolve reviews see promising signals, including gains over SAMU-XLSR in cross-modal alignment, but flag a missing acoustic-model baseline, non-significant Yemba gains in one study, single-run evaluation, and absent frugal baselines. OpenAlex data show h-index 0 and i10-index 0, 2 works in the last three years, 2 distinct coauthors, and topics clustered in speech recognition and synthesis, speech and audio processing, and graph neural networks. The field distribution splits evenly between multimodal representation learning and computational linguistics. Reproducibility scores 55/100 across 2 reviewed papers, while collaboration (22) and activity (32) reflect a small, recent, concentrated record. Output and impact currently lack sufficient data for scoring.
Collaboration pattern
Both cataloged 2026 papers are first-author works with a tight co-author set: OpenAlex records 2 distinct coauthors, and the recurring partners are Marie Tahon and Paulin Melatagia Yonta. The pattern suggests a small, stable bilateral collaboration centered on low-resource Yemba speech research rather than a broad international network; no corresponding-author role appears in the record.
Research keywords
Highlights
HGNN cross-modal knowledge transfer for Yemba word recognition
Frugal HGNN link prediction for speech-text alignment
Timeline
- 2026First-author arXiv preprint on frugal HGNN link prediction for low-resource cross-modal speech-text alignment (Yemba).
- 2026First-author arXiv preprint on HGNN cross-modal knowledge transfer for low-resource Yemba word recognition.
- 2026OpenAlex records 2 works in 2026, with topics clustered in speech recognition and synthesis and graph neural networks, and 2 distinct coauthors.
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 metrics are drawn from OpenAlex only; no ORCID is linked, so identity is resolved from public bibliographic data.
- The Evolve catalog covers 2 arXiv preprints from 2026; peer-reviewed publication status is not recorded.
- Output and impact dimensions are unscored: h-index and i10-index are 0 and no citation counts are available.
- This is an early-stage profile based on a very small record (2 works, 2 coauthors); conclusions may shift quickly as the record grows.