CSIRO researcher in computer vision: uncertainty-aware 3D reconstruction and machine learning for biodiversity
2
papers here
2
reviewed by Evolve
0
first-author
0
corresponding
—
works · OpenAlex
14
h-index · OpenAlex
Bibliometric figures come from OpenAlex (2026-09-19) 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—
The output axis is unscored: the snapshot lacks the complete per-work record this metric needs. It would measure total publication volume and its distribution across venues and years; only h-index 14 and i10-index 20 are available.
Impact—
The impact axis is unscored: citation totals are absent from this snapshot, so influence cannot be rated. It would measure citation uptake per work and standing against field baselines; h-index 14 and i10-index 20 are recorded but not converted into a score.
Focus0/100
Focus scores 0/100, the lowest of the scored axes, and the raw data supports a genuinely split portfolio: the field distribution divides exactly 50/50 between computer vision (3D reconstruction and uncertainty estimation) and machine learning for biodiversity. Topic counts likewise spread across five clusters — Advanced Vision and Imaging (16), Advanced Image Processing Techniques (11), Image Enhancement Techniques (10), AI in cancer detection (8) and Advanced Neural Network Applications (5).
Collaboration32/100
The collaboration axis scores 32/100. The recent three-year window lists only 4 distinct coauthors, and the recurring names — Lars Petersson, David Ahmedt-Aristizabal and Russell Tsuchida — indicate a tight, largely CSIRO-internal circle. In both reviewed 2026 preprints Armin appears in a middle authorship position rather than as corresponding author, consistent with a contributor role inside a small stable team.
Activity81/100
Activity is the strongest scored axis at 81/100. The snapshot records 15 works in the most recent three years, and countsByYear shows sustained volume: 5 works in 2020, a peak of 15 in 2021, then 7 in 2022, 7 in 2023 and 9 in 2024, with 5 already recorded in 2026 — including the two arXiv preprints Evolve has reviewed.
Reproducibility55/100
Reproducibility scores 55/100, computed as a mean of 55 over 2 audited papers — both rated medium reproducibility readiness. Evolve's verdicts describe careful evaluation with honest limits: GeoCond is called well evaluated though its gains are modest, while InfoTaxa rests on single-dataset evidence with approximate figure-read baselines.
Research summary
Mohammad Ali Armin is a researcher at Australia's CSIRO whose recorded portfolio splits evenly between computer vision (3D reconstruction and uncertainty estimation) and machine learning for biodiversity — a 0.5/0.5 field distribution that also drives his focus dimension score of 0. His OpenAlex-resolved snapshot (ORCID 0000-0003-4884-1557) records an h-index of 14, an i10-index of 20, and 15 works in the most recent three years, with countsByYear peaking at 15 works in 2021 and remaining steady at 7, 7 and 9 through 2022–2024. Topic frequencies centre on Advanced Vision and Imaging (16), Advanced Image Processing Techniques (11), Image Enhancement Techniques (10), AI in cancer detection (8) and Advanced Neural Network Applications (5). Evolve has reviewed two 2026 arXiv preprints. GeoCond adds a 20,000-parameter conditioning-aware reliability adapter to a frozen VGGT backbone, improving out-of-distribution AUSE from 0.32 to 0.20 and gating bundle adjustment; Evolve judged it well evaluated and honest about limits, though gains are modest. InfoTaxa proposes information-calibrated label-free clustering for fine-grained visual taxonomy on BIOSCAN-5M with BioCLIP 2; its probe-plus-DNA-barcoding audit finds species-level clustering limited by both clustering and representation quality, on single-dataset evidence. Collaboration is compact: 4 distinct recent coauthors including Lars Petersson, David Ahmedt-Aristizabal and Russell Tsuchida, with middle authorship on both reviewed papers.
Collaboration pattern
Armin's collaboration footprint in this snapshot is compact: 4 distinct recent coauthors and a collaboration dimension score of 32/100. His recurring co-authors — Lars Petersson, David Ahmedt-Aristizabal and Russell Tsuchida — point to a predominantly in-house CSIRO research group rather than a wide external network. On both Evolve-reviewed 2026 preprints he holds a middle authorship position, not corresponding author.
Research keywords
Highlights
GeoCond: a 20k-parameter reliability adapter for feed-forward 3D reconstruction
InfoTaxa: label-free clustering audited against DNA barcodes
A broad imaging portfolio with a clinical-AI strand
Beyond the two reviewed preprints, the topic profile spans Advanced Vision and Imaging (16), Advanced Image Processing Techniques (11), Image Enhancement Techniques (10), AI in cancer detection (8) and Advanced Neural Network Applications (5), alongside an h-index of 14 and an i10-index of 20.
Timeline
- 2010Earliest publication year in the OpenAlex snapshot, with 1 recorded work.
- 2020Output reached 5 recorded works, marking a step up in publication volume.
- 2021The peak year in countsByYear, with 15 recorded works.
- 20249 recorded works, the highest annual count since the 2021 peak.
- 20265 works recorded so far, including two Evolve-reviewed arXiv preprints: GeoCond on reliability-aware feed-forward 3D reconstruction and InfoTaxa on label-free clustering for fine-grained taxonomy.
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 are name-resolved from OpenAlex; h-index 14 and i10-index 20 are the only citation-level figures available in this snapshot.
- The output and impact dimensions are unscored because the underlying per-work and citation-total data are insufficient.
- Both Evolve-reviewed works are 2026 arXiv preprints with middle, non-corresponding authorship; they may not have completed peer review.
- The reproducibility score rests on only 2 audited papers, both rated medium readiness, so it should be read as indicative.