Handwriting and biometrics veteran now bridging Alzheimer's digital biomarkers and VideoLLM efficiency
2
papers here
2
reviewed by Evolve
0
first-author
0
corresponding
—
works · OpenAlex
28
h-index · OpenAlex
Bibliometric figures come from OpenAlex (2026-10-16) 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—
No score yet: the snapshot lacks the complete per-publication output ledger this axis requires. It would measure sustained publication volume and authorship roles over time; only aggregate annual counts and two recent catalog papers are available here.
Impact—
Not scored: per-paper citation data is absent from the snapshot, so influence cannot be estimated. This axis would measure citation-based reach, such as field-normalised citation rates and how often the researcher's methods are taken up elsewhere.
Focus0/100
The focus axis scores 0 because activity splits exactly evenly between two fields: computer vision (efficient multimodal machine learning) and biomedical machine learning (digital biomarkers for neurodegeneration), each at 0.5 of the field distribution. Topic counts confirm the breadth — handwritten text recognition (56), biometric identification and security (42), user authentication (39), AI in healthcare (38) and gait recognition (28) — spanning a career that predates the current two-field split.
Collaboration32/100
The collaboration axis scores 32, the weakest measured dimension. The last three years list 75 works with only 4 distinct co-authors — a thin network relative to volume. The recurring partners in the Evolve catalog are Killian Steunou, Yannis Tevissen, Changqing Gong and Huafeng Qin, and both reviewed papers place the researcher in the last author position, not corresponding, consistent with a senior-supervisor role on small teams.
Activity100/100
Activity scores 100, the axis maximum. The record lists 75 works in the last three years, and annual counts of 25 (2023), 35 (2024), 21 (2025) and 19 (2026 to date); the 2024 figure is the highest in a publishing trail reaching back to 1996 (1 work) and includes earlier peaks of 8 in 2000 and 24 in 2017.
Reproducibility38/100
Reproducibility scores 38, the mean across the 2 Evolve-reviewed papers. The online-handwriting Alzheimer's paper is rated low readiness: the review credits solid engineering (92.55% BAcc on DARWIN) but notes missing significance tests and tuning protocols. The VideoLLM efficiency survey is rated medium, with disciplined comparisons whose evidence rests on borrowed re-runs rather than new experiments.
Research summary
Mounîm A. El-Yacoubi (Institut Mines-Télécom; ORCID 0000-0002-7383-0588) is a senior pattern-recognition researcher whose bibliometric record spans three decades, from the first entries in 1996 to an intense recent run. The snapshot credits an h-index of 28 and an i10-index of 71, with 75 works in the last three years and a peak of 35 publications in 2024 — the basis for a maximal activity score of 100. Topical mass sits in handwritten text recognition (56 topic hits), biometric identification and security (42), user authentication (39), AI in healthcare (38) and gait recognition (28). Current work splits evenly between two fields per the field distribution: efficient multimodal machine learning in computer vision and biomedical machine learning for neurodegeneration digital biomarkers, which drives the focus score of 0. Two 2026 arXiv papers in the Evolve catalog — a weakly supervised online-handwriting system for Alzheimer's detection (92.55% BAcc on DARWIN, per the review) and a survey of inference-efficiency mechanisms across 125 VideoLLM works — both list the researcher as last author, not corresponding. Evolve rates reproducibility at 38 (mean of two papers), flagging absent significance tests in the first and reliance on borrowed re-runs in the second. Collaboration scores 32: 75 recent works with only 4 distinct co-authors, with Steunou, Tevissen, Gong and Qin recurring.
Collaboration pattern
Collaboration is narrow but stable: 75 works over the last three years with only 4 distinct co-authors (collaboration score 32). The recurring partners are Killian Steunou, Yannis Tevissen, Changqing Gong and Huafeng Qin, and on both 2026 catalog papers the researcher appears as last author without corresponding status — a small-team, senior-supervisor pattern rather than large-consortium work.
Research keywords
Highlights
Weakly supervised handwriting risk discovery for Alzheimer's detection
A stage-by-stage map of VideoLLM inference efficiency
A three-decade record with a late-career output surge
The bibliometric trail runs from 1996 to 2026 (h-index 28, i10-index 71), with the last three years contributing 75 works and 2024 alone 35 — the busiest single year on record.
Timeline
- 1996Earliest entries in the bibliometric record: 1 publication in 1996 and 2 in 1997, opening a three-decade publishing trail.
- 2000An early output peak of 8 publications in 2000, repeated in 2002, during the handwriting and biometrics era.
- 2017Output climbs to 24 publications — the highest annual count of the 2010s in the record.
- 2023Recent acceleration begins: 25 publications in 2023, on the way to 75 works across the last three years.
- 2024A record 35 publications in 2024, the busiest single year in the snapshot.
- 2026Two arXiv-first papers enter the Evolve catalog — online-handwriting Alzheimer's detection and a VideoLLM efficiency survey — both with the researcher as last author.
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.
- Profile is name-resolved from public bibliographic data (OpenAlex identity), anchored to ORCID 0000-0002-7383-0588.
- Bibliometrics come from OpenAlex only; per-paper citation counts are absent, so the output and impact axes could not be scored.
- The reproducibility mean (38) rests on just two Evolve-reviewed papers, both 2026 arXiv preprints.
- The record lists country code 'CN' while the affiliation shown is Institut Mines-Télécom; this mismatch was not verified.