Nanjing University computer vision researcher in instruction-based video editing and embodied-agent spatial memory
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—
Reliable publication-volume aggregates are not yet available in this profile's OpenAlex record. This axis would measure total works, preprint versus venue output, and yearly publication distribution; it stays unscored rather than estimated from partial data.
Impact—
Citation indicators here are absent or uninformative: the record reports h-index and i10-index as zero, which reflects missing citation coverage rather than measured impact. This axis would capture citation counts, citing works, and influence on follow-up research.
Focus0/100
Focus scores 0, the lowest axis. The field distribution splits evenly (0.5 and 0.5) between video generation/editing and embodied AI with long-horizon video understanding, and the topic profile spans five clusters—from GAN-based image synthesis (9) to face recognition (4) and 3D shape modeling (3)—indicating broad, unsettled topical concentration.
Collaboration60/100
Collaboration scores 60. The OpenAlex record lists 19 distinct co-authors, with 12 recurring collaborators including Tianyidan Xie, Shenyi Wang, and Jian Yang. On both reviewed 2026 papers Zili Yi sits in the last author position without a corresponding-author flag, a pattern consistent with a senior supervisory role within a moderately sized, stable team rather than a lone-author or large consortium workflow.
Activity75/100
Activity scores 75, the strongest axis in this profile. The record shows 12 works within the recent three-year window, and the yearly counts place all 12 in 2026. Combined with two arXiv preprints reviewed by Evolve in 2026, this indicates a burst of recent output concentrated in the current year rather than a long cumulative record.
Reproducibility20/100
Reproducibility scores 20, based on Evolve's review of 2 papers (reproMean 20). Both reviewed works are rated low readiness: VideoX-Qwen reports no ablations and small evaluations, and the spatial-memory paper states no code release and a tracker-coupled benchmark. Public artifacts would need to accompany future work to raise this score.
Research summary
Zili Yi is a computer vision researcher at Nanjing University whose recent output, as recorded by OpenAlex, splits evenly between two subfields: video generation and editing, and embodied AI with long-horizon video understanding. The topic profile spans five clusters, led by generative adversarial networks and image synthesis (9), followed by multimodal machine learning applications (5), face recognition and analysis (4), 3D shape modeling and analysis (3), and image enhancement techniques (3). Two 2026 arXiv preprints reviewed by Evolve anchor this profile. VideoX-Qwen presents a data-centric, instruction-based video editing framework built on a 1.2M-record pipeline and a Qwen-Wan editor that reportedly wins 9 of 11 metrics; Evolve credits solid engineering while flagging a small evaluation, judge bias, and missing ablations. The second preprint introduces per-object linguistic trajectory encoding for spatial memory in embodied agents plus a new SMB benchmark, with reported gains of +13.4 and +14.3 SMB, though without significance tests or stated code. Bibliometrically, the record shows 12 works in 2026, covering the entire recent three-year window, 19 distinct co-authors, and 12 recurring collaborators. Zili Yi appears as last author on both reviewed papers without a corresponding-author flag, a position consistent with a senior or supervisory role. Citation aggregates are uninformative in the current record, and both reviewed papers carry low reproducibility readiness, making impact and reproducibility the weakest axes.
Collaboration pattern
Zili Yi's OpenAlex record lists 19 distinct co-authors and a recurring core of 12, including Tianyidan Xie, Shenyi Wang, Qiang Tang, Mingjie Wang, Jian Yang, and Zhan Xu. On both 2026 papers reviewed by Evolve, Zili Yi is the last author without a corresponding-author flag, consistent with a senior or supervisory role. The footprint is a moderate-sized, stable team centered at Nanjing University, with output split across video-editing and embodied-AI lines.
Research keywords
Highlights
VideoX-Qwen: data-centric instruction-based video editing
Linguistic trajectory encoding for embodied spatial memory
Timeline
- 2026Evolve reviews the VideoX-Qwen preprint (arXiv, 2026) on data-centric instruction-based video editing, with Zili Yi in the last author position.
- 2026Evolve reviews the linguistic trajectory encoding preprint on embodied-agent spatial memory, again listing Zili Yi as last author.
- 2026OpenAlex records 12 works for 2026, covering the entire recent three-year window, alongside 19 distinct co-authors.
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.
- Bibliometric indicators are computed from OpenAlex only; no ORCID is linked, so this profile is name-resolved from public bibliographic data.
- The record reports h-index and i10-index as zero, so the output and impact dimensions are not scored.
- Only 2 papers have been independently reviewed by Evolve, and both are 2026 arXiv preprints with low reproducibility readiness.
- All 12 recorded works fall in 2026, so longer-term productivity and citation trends cannot be assessed from this snapshot.