Task-aware 4D radar preprocessing for autonomous driving — KAIST researcher with GNSS and inertial navigation roots
5
papers here
4
reviewed by Evolve
0
first-author
0
corresponding
—
works · OpenAlex
22
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 yet scored. The snapshot includes raw counters (h-index 22, i10-index 39), but no output-volume axis has been computed for this profile. This dimension would measure total publication volume and its distribution across years and venues.
Impact—
Not yet scored. Citation-based fields are absent from the computed radar, so no influence estimate is offered here. This dimension would measure citation counts, field-normalized citation impact, and the scholarly reach of each work.
Focus0/100
Scores 0 on focus, reflecting strong topic dispersion: five clusters are nearly flat — autonomous vehicle technology and safety (37), advanced neural network applications (36), GNSS positioning and interference (34), radar systems and signal processing (30), and inertial sensor and navigation (30). Within the Evolve catalog itself, the field distribution is concentrated on autonomous-driving perception with 4D radar (0.5/0.5 across two labels of the same field).
Collaboration48/100
Scores 39 on collaboration. The recent three-year window pairs 45 works with only 6 distinct coauthors, a compact working group; the recurring partners are Dong-Hee Paek, Seung-Hyun Song, Woo-Jin Jung, and Jeong-Su Park. Both Evolve-catalog papers place Kong in the last author position without corresponding status.
Activity100/100
Scores 100 on activity. The counts-by-year series shows 10 works in 2024, 15 in 2025, and 20 in 2026 — a career-high climb — with 45 works inside the most recent three-year window. This is the profile's strongest axis and indicates a steeply accelerating publishing pace.
Reproducibility55/100
Scores 55 (reproMean) across 2 Evolve-reviewed papers, both tagged medium reproducibility readiness. The verdicts flag concrete gaps: STAR reports +5.6 AP_BEV on K-Radar but on a single dataset without preprocessing baselines or variance data, and the preprocessing study shows FPS costs, omits ablations, and rests its embedded claims on RTX 3090 runs.
Research summary
Seung-Hyun Kong is a researcher at the Korea Advanced Institute of Science and Technology (KAIST), South Korea, whose record spans GNSS positioning and interference, radar systems and signal processing, inertial sensor navigation, and — most recently — 4D imaging radar for autonomous-driving perception. The OpenAlex snapshot reports an h-index of 22 and an i10-index of 39, with a publication series beginning in 2007 and accelerating sharply: 10 works in 2024, 15 in 2025, and 20 in 2026, totaling 45 works over the last three years. Topic clusters are led by autonomous vehicle technology and safety (37), advanced neural network applications (36), and GNSS positioning and interference (34), tracing a move from classical navigation toward learning-based perception. Both Evolve-reviewed papers are 2026 arXiv preprints with Kong in the last author position: STAR, a scene- and task-aware 4D radar preprocessing network that gains +5.6 AP_BEV on K-Radar, and an accuracy- and real-time-aware preprocessing study introducing P3DP, MF-KDE, and an ENS metric. Evolve scores activity at 100, reproducibility at 55, collaboration at 39, and focus at 0; the output and impact axes are not yet computed, a limitation carried into the caveats.
Collaboration pattern
Collaboration is concentrated in a small, stable group: the recent three-year window pairs 45 works with only 6 distinct coauthors (collaboration score 39). The recurring partners are Dong-Hee Paek, Seung-Hyun Song, Woo-Jin Jung, and Jeong-Su Park. On both Evolve-catalog papers Kong sits in the last author position without corresponding status, consistent with a senior supervisory role over a compact radar-perception group rather than a wide external network.
Research keywords
Highlights
STAR: task-aware 4D radar preprocessing gains +5.6 AP_BEV on K-Radar
Accuracy- and real-time-aware radar preprocessing: P3DP, MF-KDE, and an ENS metric
From GNSS and inertial navigation to cognitive radar
Topic counts position autonomous vehicle technology and safety (37) and advanced neural network applications (36) alongside GNSS positioning and interference (34), radar systems and signal processing (30), and inertial sensor and navigation (30) — a research line that extends long-standing navigation expertise into learning-based radar perception.
Timeline
- 2007Earliest recorded publication year in the OpenAlex counts-by-year series (1 work).
- 2014–2015Output climbs to 9 works in 2014 and 14 in 2015, an early peak in the record.
- 2022A second peak: 13 works recorded in 2022.
- 202515 works recorded — the highest annual output to that point.
- 202620 works recorded to date, a new high; includes both Evolve-reviewed arXiv papers on 4D radar preprocessing (STAR and the P3DP/MF-KDE study), each with Kong as last author.
Papers in the Evolve catalog
STAR: Scene- and Task-Aware 4D Radar Preprocessing Towards End-to-End Cognitive Radar
Accuracy- and Real-Time-Aware 4D Radar Preprocessing for Autonomous Driving Perception Systems
IMM-based Multiple Object Tracking using a State Prediction Neural Network
Frequency Estimation Based on SNR-adaptive Frequency Estimator Under Wide SNR Range
DNC-IMM: Early Lane-Change Intention Recognition via Neural Calibration Based on Driving Context Information
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.
- Bibliometric indicators and topic counts derive from an OpenAlex snapshot; venue and citation coverage may be incomplete.
- Output and impact dimensions are unscored in this profile, so no claim is made about total output volume or citation influence beyond the raw h-index (22) and i10-index (39).
- The reproducibility axis rests on only 2 Evolve-reviewed papers, both 2026 arXiv preprints with medium readiness.
- Identity is name-resolved from public bibliographic data (ORCID 0000-0002-4753-1998, OpenAlex); no biographical information beyond these sources is included.