Yong Man Ro

Korea Advanced Institute of Science and Technology · South Korea

Unclaimed · AI-structured profileOpenAlex

KAIST researcher spanning speech/multimodal ML and remote-sensing multimodal LLMs; last author on 2026 arXiv preprints.

papers here

reviewed by Evolve

first-author

corresponding

—

works · OpenAlex

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—

Not scored yet: this OpenAlex identity lists only sparse works (1 in 2003, 1 in 2024, 12 in 2026), so a reliable total publication count cannot be aggregated. This axis would measure total and per-year publication volume across venues.

Impact—

Impact is not scored: citation data for this identity are too thin (recorded h-index 1, i10-index 0) to support a fair estimate. This axis would measure citations, h-index trajectory, and downstream scholarly influence of the work.

Focus0/100

Scores 0/100, the widest topical spread in this snapshot: the field distribution splits evenly at 0.5 and 0.5 between speech and multimodal machine learning and computer vision for remote sensing multimodal LLMs, while OpenAlex topics scatter across five clusters — multimodal machine learning applications (10), topic modeling (4), advanced neural network applications (3), speech recognition and synthesis (3), and visual attention and saliency detection (3).

Collaboration44/100

Scores 44/100. The snapshot records 8 distinct coauthors in the recent window, with a recurring core group — Sungjune Park, Yeonju Kim, Yeongyun Kim, Juwan Chung, Youngchae Chee, Hosu Lee, Junho Kim, and Bella Godiva — and all three catalog papers place this researcher in the last author position, a pattern consistent with senior or supervisory authorship within a compact KAIST-led team rather than large multi-institution consortia.

Activity77/100

Scores 77/100 on sustained recent output: 13 works in the last three years, with counts by year peaking at 12 works in 2026 versus 1 in 2024 and 1 in 2003. All three Evolve-catalog papers are 2026 arXiv preprints, indicating a concentrated, currently intense publication push across remote sensing vision-language and speech-dialogue problems.

Reproducibility55/100

Scores 55/100 (reproMean 55 across 2 scored papers). Evolve reviews rate both catalog entries at medium readiness: the change-localization paper reports large controlled gains on xBD/S2Looking but smaller ones on HRCUS/LEVIR with no significance tests, and AV-STE shows consistent gains and good ablations yet evaluates on only 125 clips, also without significance testing.

Research summary

Yong Man Ro is a researcher at the Korea Advanced Institute of Science and Technology (KAIST), South Korea; this profile is name-resolved from public bibliographic data (an OpenAlex identity with no linked ORCID). The snapshot describes an evenly split research program: 0.5 of the field distribution in speech and multimodal machine learning and 0.5 in computer vision for remote sensing multimodal LLMs. OpenAlex topic counts cluster around multimodal machine learning applications (10), topic modeling (4), advanced neural network applications (3), speech recognition and synthesis (3), and visual attention and saliency detection (3). Recent activity is strong: 13 works in the last three years, 12 of them dated 2026. The Evolve catalog holds three 2026 arXiv preprints, each with this researcher in the last author position. Two carry Evolve reviews: a region-selection approach to temporal change localization in remote sensing multimodal LLMs, with large controlled gains on xBD/S2Looking but smaller ones on HRCUS/LEVIR, and AV-STE, a noise-adaptive audio-visual speech token enhancement method for full-duplex spoken dialogue over a frozen Moshi model, evaluated on a 125-clip set. A third, PRISM, targets view-invariant video representation learning. Citation-based output and impact dimensions are not scored because the underlying OpenAlex identity is sparsely populated (recorded h-index 1, i10-index 0).

Collaboration pattern

The snapshot records 8 distinct coauthors in the recent window, with a recurring core group — Sungjune Park, Yeonju Kim, Yeongyun Kim, Juwan Chung, Youngchae Chee, Hosu Lee, Junho Kim, plus Bella Godiva. All three Evolve-catalog papers place this researcher in the last author position (none flagged corresponding), consistent with a supervisory role within a compact KAIST-led team rather than large multi-institution consortia. Collaboration dimension score: 44/100.

Research keywords

Multimodal machine learningMultimodal large language modelsRemote sensing change detectionTemporal change localizationVisual groundingAudio-visual speech processingFull-duplex spoken dialogueSemantic speech tokensLip readingNoise robustnessVideo representation learningVisual attention and saliency

Highlights

Region selection replaces coordinate generation for temporal change localization
This 2026 arXiv paper swaps coordinate generation for candidate region selection in remote sensing multimodal LLMs; Evolve's review notes large controlled gains on xBD/S2Looking but small ones on HRCUS/LEVIR, with no significance tests reported.
Noise-adaptive audio-visual speech token enhancement for full-duplex dialogue
AV-STE uses lip video to restore noisy speech tokens for a frozen Moshi model; Evolve's review cites consistent gains and good ablations, but no significance tests and a 125-clip evaluation set.
PRISM: view-invariant video representation via semantic latent decomposition
A third 2026 arXiv preprint in the catalog proposes predictive recomposition via semantic latent decomposition for view-invariant video representations; it has not yet received a full Evolve review, so no findings are reported here.

Timeline

2003Earliest work in the OpenAlex snapshot: 1 publication recorded for this year.
2024Counts-by-year data record 1 work for this year.
2026Most productive year in the snapshot: 12 works recorded; three arXiv preprints enter the Evolve catalog — region-selection change localization, AV-STE for full-duplex spoken dialogue, and PRISM for video representation learning — each with last-author placement.

Papers in the Evolve catalog

Frequent co-authors

Sungjune Park×2 Bella Godiva×1 Yeonju Kim×1 Juwan Chung×1 Yeongyun Kim×1 Youngchae Chee×1 Hosu Lee×1 Junho Kim×1

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 identity (identity: openalex); no ORCID is linked, and the sparsely populated record (h-index 1, i10-index 0; 1 work in 2003 and 1 in 2024) suggests partial coverage of this name.
Output and impact dimensions are unscored because citation aggregation for this identity is insufficient (radar dimMissing); no citation counts beyond the recorded h/i10 indices are available.
Paper-level evidence covers three 2026 arXiv preprints, two of which carry Evolve verdicts; all claims rest on those reviews and the provided metadata only.
The focus score of 0 reflects the even 0.5/0.5 split across two distinct fields in this snapshot; it describes the recorded record, not a judgment of research quality.