Early-stage 3D point cloud tracking researcher — co-author of SAVTrack and MAETrack for LiDAR single object tracking
2
papers here
2
reviewed by Evolve
0
first-author
0
corresponding
—
works · OpenAlex
0
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—
Not yet scored: this OpenAlex record is too sparse to gauge total publications and recent productivity reliably. Only two 2026 preprints are currently linked, so no output score is estimated.
Impact—
Not yet scored: citation and h-index data are effectively empty (reported h-index 0) in this OpenAlex snapshot, so the impact axis — which measures citation-based influence of published work — cannot be reliably estimated.
Focus0/100
Focus scores 0, the lowest value, because the record is too young to show a concentrated topical signature: OpenAlex spreads this name across five topics (Video Surveillance and Tracking Methods counted twice; four others once each). The two reviewed papers, however, both sit in 3D point cloud tracking, with fieldDist split 50/50 between two closely related subfields.
Collaboration39/100
Collaboration scores 39. The snapshot records 6 distinct coauthors, and the frequent-coauthor list is short — Sifan Zhou, Linyue Tan, Xiaobo Lu, Qiwei Wang and Ziyu Liu. On both reviewed papers Ziyu Zhao appears in a middle author position without corresponding-author role, consistent with contributing within a small, stable team rather than leading or spanning large collaborations.
Activity32/100
Activity scores 32, reflecting a thin recent footprint: the counts-by-year series contains only 2026 with 2 works, and recent3yWorks is 2. Both records are arXiv preprints from 2026, so measured output is concentrated in a single year and the snapshot likely understates ongoing research activity.
Reproducibility55/100
Reproducibility scores 55, averaged over 2 papers. Both reviewed papers carry a medium reproducibility-readiness rating: SAVTrack reports consistent ~2-3 point KITTI/nuScenes gains at 82 FPS, while MAETrack offers strong ablations but retains single-run statistics and a pretraining-data confound, per Evolve's review.
Research summary
Ziyu Zhao is an early-stage computer vision researcher whose traceable work centers on 3D single object tracking in LiDAR point clouds, a core perception task for autonomous driving. This profile is name-resolved from public OpenAlex bibliographic data; no institution, country, or ORCID is attached, and the record itself is sparse — h-index 0, i10-index 0, and a counts-by-year series containing only 2026 with 2 works. The substantive evidence comes from two 2026 arXiv preprints reviewed by Evolve. SAVTrack introduces selective vote aggregation with motion-conditioned vote gating on top of a P2P baseline, delivering consistent ~2-3 point gains on KITTI and nuScenes at 82 FPS — assessed as solid but incremental. MAETrack transfers masked-autoencoder pretrained geometric priors to 3D tracking via shallow-layer initialization and search-region gating, with strong ablations but lingering single-run statistics and a pretraining-data confound. Both papers place Ziyu Zhao in a middle author position without corresponding role, within a compact coauthor set of 6 distinct collaborators, including recurring names such as Sifan Zhou, Linyue Tan, Xiaobo Lu, Qiwei Wang, and Ziyu Liu. Radar scores are collaboration 39, activity 32, reproducibility 55, and focus 0, with output and impact axes missing. OpenAlex topics scatter across video surveillance and tracking, neural network applications, 3D shape modeling, human pose, and face analysis, but the reviewed work sits squarely in LiDAR point cloud tracking.
Collaboration pattern
The record shows a small-team collaboration pattern: 6 distinct coauthors overall, with a recurring core of Sifan Zhou, Linyue Tan, Xiaobo Lu, Qiwei Wang and Ziyu Liu. On both 2026 arXiv papers Ziyu Zhao sits in a middle author position and is not the corresponding author, suggesting a contributor role within a stable 3D tracking group rather than independent or cross-institution leadership. Collaboration score: 39/100.
Research keywords
Highlights
Selective vote aggregation for reliability-aware point cloud tracking
Pretrained geometric priors via shallow-layer MAE initialization for 3D LiDAR tracking
Timeline
- 2026Co-authored SAVTrack, proposing selective vote aggregation with motion-conditioned gating for reliability-aware 3D point cloud tracking, evaluated on KITTI/nuScenes (arXiv preprint).
- 2026Co-authored MAETrack, transferring masked-autoencoder pretrained geometric priors to 3D single object tracking through shallow-layer initialization and search gating (arXiv preprint).
- 2026OpenAlex snapshot records 2 works in this single year, both in LiDAR point cloud tracking for autonomous driving perception.
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.
- Identity is name-resolved from public OpenAlex bibliographic data; no institution, country, or ORCID is attached to this record.
- All bibliometrics come from an OpenAlex snapshot only; the reported h-index (0) and i10-index (0) indicate a sparse record, not necessarily a complete publication history.
- Qualitative assessments rest on only 2 papers (both 2026 arXiv preprints) reviewed by Evolve; output and impact scores are missing for this reason.
- Frequent-coauthor names are listed as published strings and the list contains a duplicate entry (Qiwei Wang appears twice).