Seung-Kyum Choi
Georgia Institute of Technology · United States
Georgia Tech researcher pairing machine-learning vibration diagnosis with CAD-free 3D industrial perception
2
papers here
2
reviewed by Evolve
0
first-author
0
corresponding
—
works · OpenAlex
0
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 output score is available: the OpenAlex record underlying this profile carries only four works, all dated 2026, without aggregated per-year publication totals. This axis would normally measure total paper count, annual output trend, and venue distribution.
Impact—
Impact cannot be scored yet: the snapshot reports h-index 0 and i10-index 0, typical of a newly indexed record, so citation-based standing is uninformative. This axis would measure citation counts, h-index, and field-normalized percentile.
Focus0/100
Focus scores 0 of 100, reflecting maximal topical spread rather than depth in one lane. The field distribution splits exactly 50/50 between machine learning for vibration-based condition monitoring and computer vision / robotics perception for industrial object recognition, and topic tags scatter across five clusters (machine fault diagnosis 2; advanced neural network applications, reinforcement learning in robotics, structural health monitoring, and robotic path planning 1 each). The two cataloged 2026 papers anchor these distinct tracks.
Collaboration22/100
Collaboration scores 22 of 100, grounded in just 2 distinct coauthors across the OpenAlex record. The named recurring partners are Mainak Mallick and Chenxi Tao, and on both cataloged 2026 papers Seung-Kyum Choi is listed in the last author position — a small-team, senior-author pattern rather than large-consortium work, with no cross-institution links visible in the data.
Activity47/100
Activity scores 47 of 100, mid-range and heavily time-concentrated: the counts-by-year series contains a single entry, 4 works in 2026, and recent3yWorks is likewise 4. All indexed output therefore falls inside the current year, which lifts recent output but cannot yet demonstrate sustained multi-year productivity. A longer indexed history would be needed to judge whether this pace persists.
Reproducibility55/100
Reproducibility scores 55 of 100, averaged over the 2 reviewed papers, each rated medium readiness. Evolve's verdicts flag concrete gaps: the vibration-diagnosis paper offers broad, honest evaluation but thin significance testing and ablation rigor, while the 3D-shape-prior study shows solid internal evidence yet covers only 56 objects with no external baselines. A mid-range score fairly captures careful methods paired with incomplete validation.
Research summary
Seung-Kyum Choi is a researcher at Georgia Institute of Technology whose profile, built from an OpenAlex bibliographic record, is best described as early-stage: the indexed snapshot lists 4 works, all dated 2026, with h-index 0 and i10-index 0. Evolve has reviewed two 2026 arXiv papers, on both of which Choi appears in the last author position. They define two parallel research tracks. The first, in machine learning for vibration-based condition monitoring, contributes a 40k-parameter oscillatory state space model for fault diagnosis when labels are scarce, with broad and honest evaluation albeit thin significance testing and ablation rigor. The second, in computer vision and robotics perception, introduces a scan-derived 3D Gaussian Splatting geometry prior that lets frozen foundation features such as DINOv2 recognize industrial objects without CAD models, validated on 56 objects with solid internal evidence but no external baselines. Topic tags span five clusters — machine fault diagnosis (2), advanced neural network applications, reinforcement learning in robotics, structural health monitoring, and robotic path planning — matching an even 50/50 field split between the two tracks. Collaboration is narrow: 2 distinct coauthors, with Mainak Mallick and Chenxi Tao the recurring partners. Both reviewed papers carry medium reproducibility readiness, averaging to a reproducibility score of 55. Independent verification of impact must wait until citation data accumulates.
Collaboration pattern
The OpenAlex record shows a deliberately small collaboration footprint: 2 distinct coauthors, both recurring — Mainak Mallick and Chenxi Tao. On both 2026 cataloged arXiv papers Choi occupies the last author position without corresponding-author designation, a pattern consistent with supervising small project teams. No cross-institution or large-consortium links appear in the data, and the collaboration dimension score of 22 reflects that narrow, stable circle.
Research keywords
Highlights
Oscillatory state space model for scarce-label vibration diagnosis
CAD-free 3D geometry prior that complements vision foundation models
Timeline
- 2026Posts an oscillatory state space model of about 40k parameters for vibration fault diagnosis under scarce labels on arXiv, listed in the last author position.
- 2026Publishes a CAD-free study in which a scan-derived 3D Gaussian Splatting shape prior complements frozen vision foundation models such as DINOv2, again as last author.
- 2026OpenAlex records 4 works for this year (recent3yWorks also 4) — the entirety of the indexed output to date.
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 come from a single OpenAlex snapshot; no ORCID is linked, so identity is name-resolved from public bibliographic data and disambiguation from same-name researchers is not guaranteed.
- The indexed record is fresh: all 4 works are dated 2026 and h-index / i10-index are 0, so citation-based impact cannot be assessed.
- Only the two 2026 arXiv preprints have been reviewed by Evolve; any earlier publications fall outside the reviewed scope of this profile.
- Reproducibility and readiness judgments rest on 2 papers rated medium each; the sample is too small to generalize.