Researcher bridging few-shot computer vision and thermal-fluid engineering, Norwegian University of Life Sciences
3
papers here
1
reviewed by Evolve
0
first-author
0
corresponding
—
works · OpenAlex
7
h-index · OpenAlex
Bibliometric figures come from OpenAlex (2026-10-17) 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 scoring is not yet available. This axis would measure annual and cumulative publication volume against field norms; the snapshot lists per-year work counts (peaking at 5 in 2022 and 2023, 4 in 2026) but no normalized output score.
Impact—
Citation-impact scoring is not available: this axis would capture citation counts, h-index percentile placement, and citing-work quality. Only raw indicators exist — h-index 7 and i10-index 7 — so no impact score is reported.
Focus100/100
The maximum focus score of 100 reflects a tightly concentrated record: five OpenAlex topic clusters cover essentially two strands — learning-based vision (Domain Adaptation and Few-Shot Learning, 6 works; Advanced Neural Network Applications, 5) and thermal engineering (Solar Thermal and Photovoltaic Systems, 6; Heat Transfer Mechanisms, 4; Nanofluid Flow and Heat Transfer, 3) — with no scattered long tail across unrelated fields.
Collaboration50/100
A mid-range collaboration score of 50 matches a modest network: 11 distinct coauthors over the recent three-year window containing 6 works. Recurring partners include Fadi Al Machot, Chandler Timm C. Doloriel, and Habib Ullah. In all three catalogued 2026 papers he appears in a middle author position without corresponding status, consistent with team-based rather than lead roles.
Activity57/100
An activity score of 57 reflects moderate, uneven cadence: yearly counts run 1 (2019), 2 (2020), 5 (2022), 5 (2023), 1 (2024), 1 (2025), and 4 (2026), with 6 works in the last three years. After the 2022–2023 peak the pace thinned in 2024–2025 before rebounding in 2026, largely through arXiv preprints.
Reproducibility55/100
Reproducibility is scored at 55 (reproMean) from just one deeply reviewed paper, so the signal is thin. Evolve's review of the 2026 meta-learning plant growth paper praises honest ablations and solid statistics, flags random-split data-leakage risk, and rates reproduction readiness as medium.
Research summary
Muhammad Salman Siddiqui is a researcher at the Norwegian University of Life Sciences whose bibliometric record, resolved from OpenAlex, spans two coherent strands: learning-based computer vision and thermal-fluid engineering. OpenAlex topic modeling shows this directly — Domain Adaptation and Few-Shot Learning (6 works) and Advanced Neural Network Applications (5) sit alongside Solar Thermal and Photovoltaic Systems (6), Heat Transfer Mechanisms (4), and Nanofluid Flow and Heat Transfer (3). The 2026 catalog entries are all computer vision: a few-shot meta-learning paper estimating plant growth with vision transformers and fuzzy c-means task construction, and two preprints on continual test-time adaptation, one targeting vision-language models in open-vocabulary semantic segmentation and the other an entropy sensitivity-guidance method in a strict online setting. Raw indicators remain at early-career scale: h-index 7 and i10-index 7, with yearly outputs peaking at five in 2022 and 2023, and 6 works in the last three years alongside 11 distinct coauthors. Evolve's review of the plant growth paper praises honest ablations and solid statistics while flagging random-split leakage risk, rating reproduction readiness medium. In all three catalogued papers he is a middle author and not corresponding, suggesting contributor rather than lead roles. Radar scores place focus at 100, activity at 57, collaboration at 50, and reproducibility at 55; output and impact are not yet scored.
Collaboration pattern
With 11 distinct coauthors across 6 recent works, collaboration is modest and team-based. Recurring names include Fadi Al Machot, Chandler Timm C. Doloriel, and Habib Ullah, with Kristian Hovde Liland and Tor Kristian Stevik also among frequent partners. In all three 2026 catalog papers he appears as a middle author and not corresponding, suggesting contributor roles within larger machine learning teams rather than leading his own group.
Research keywords
Highlights
Meta-learning for data-efficient plant growth estimation
Continual test-time adaptation of vision-language models in open-vocabulary segmentation
A second 2026 arXiv preprint extends continual test-time adaptation to vision-language models in open-vocabulary semantic segmentation, pushing adaptation beyond closed classification settings. It is not yet deeply reviewed by Evolve, so no verdict or reproduction rating is available.
Entropy sensitivity-guidance for strict online continual adaptation
A third 2026 arXiv preprint addresses continual test-time adaptation via entropy sensitivity-guidance in a strict online setting, where updates must rely on unlabeled test data as they arrive. Evolve has not yet issued a review verdict or reproduction rating for this work.
Timeline
- 2019First logged output appears in the bibliometric record (1 work in 2019), the start of a profile spanning thermal engineering and machine learning.
- 2022Output accelerates to 5 works in 2022, with topics spanning solar thermal and photovoltaic systems, nanofluid flow, and neural network applications.
- 2023A second peak of 5 works in 2023 consolidates the two research strands of learning-based vision and heat transfer.
- 2024Pace thins to a single logged work in 2024, the start of a quieter two-year stretch.
- 2026Three arXiv preprints appear in 2026, led by a deeply reviewed meta-learning paper on few-shot plant growth estimation with vision transformers, plus two works on continual test-time adaptation.
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 (h-index, i10-index, yearly counts, topics, coauthors) come from an OpenAlex identity snapshot; the profile is name-resolved from public bibliographic data and may miss or merge works.
- Output and impact radar scores are unavailable; only raw indicators (h-index 7, i10-index 7) are reported.
- The reproducibility score (55) rests on a single deeply reviewed paper, so it is a thin signal.
- Two of the three catalogued 2026 papers have not yet been reviewed by Evolve; in all catalogued papers the author position is middle author and not corresponding.