Kanagawa University researcher in zero-shot anomaly detection and implicit neural video compression
2
papers here
2
reviewed by Evolve
0
first-author
0
corresponding
—
works · OpenAlex
3
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—
Publication-output metrics for this profile are not yet computed: the OpenAlex snapshot lacks a consolidated work count and volume percentile for this author. This axis would measure total publication volume and its trajectory across the career.
Impact—
Only minimal citation signals exist (h-index 3, i10-index 3 in the OpenAlex snapshot) with no citation totals or field-normalized percentiles, so impact is not scored. This axis would measure citation influence against field baselines.
Focus0/100
The focus axis scores 0. The underlying OpenAlex field distribution splits evenly — computer-vision anomaly detection at 0.5 and neural video compression at 0.5 — so no single specialization dominates, and the topic profile spreads further across five clusters such as anomaly detection techniques (5) and digital media forensic detection (4).
Collaboration41/100
Collaboration scores 41, a moderate level. The OpenAlex record shows 7 distinct coauthors over the last three years alongside 7 recent works, with a recurring group of Xuezhi Xiang, Heqi Xiang, Jiayao Liu, Guanghao Wu, Xiaoheng Li, Yiming Chen and Yixin Zhao. Both cataloged 2026 papers place Zhang in the last author position without corresponding-author status.
Activity61/100
Activity scores 61. Recent output is steady and rising: the OpenAlex counts-by-year series shows 1 work in 2022 and 2023, 2 in 2024 and 2025, and 3 in 2026, with 7 works in the last three years, including two 2026 arXiv preprints spanning anomaly detection and neural video compression.
Reproducibility55/100
Reproducibility scores 55, computed from Evolve's review of 2 papers (mean readiness 55). Both are rated medium: PSMP-CLIP reports ablated pixel-level gains on 7 of 11 benchmarks but omits key hyperparameters and code, while TCNeRV reports +2.2 dB over HNeRV-Boost on UVG at 3M parameters but validates on a single dataset and fades at 10M scale.
Research summary
Shanjun Zhang is a researcher at Kanagawa University, Japan, whose recent work concentrates on two computer-vision tracks: zero-shot anomaly detection with vision-language models, and implicit-neural-representation video compression. The OpenAlex snapshot records 7 works in the last three years, an h-index of 3 (i10-index 3), and annual counts rising from 1 work in 2022 to 3 in 2026. The field distribution splits evenly between anomaly detection (0.5) and neural video compression (0.5). Two 2026 arXiv papers are in the Evolve catalog: PSMP-CLIP, a CLIP+SAM2 zero-shot anomaly detector with ablated pixel-level gains on 7 of 11 benchmarks, though image-level performance lags rivals; and TCNeRV, which adds dual-domain temporal context modeling to INR video coding, reporting +2.2 dB over HNeRV-Boost on UVG at 3M parameters. Both papers list Zhang as last author without corresponding status, consistent with a senior role in a stable seven-person coauthor group. Evolve rates both medium reproducibility (mean 55): PSMP-CLIP omits key hyperparameters and code, while TCNeRV is validated on a single dataset.
Collaboration pattern
Collaboration centers on a stable seven-person coauthor group — Xuezhi Xiang, Heqi Xiang, Jiayao Liu, Guanghao Wu, Xiaoheng Li, Yiming Chen and Yixin Zhao — with 7 distinct coauthors recorded over the last three years. On both cataloged 2026 arXiv papers Zhang appears as last author without corresponding-author status, a pattern typical of a senior or supervisory role within a continuing team rather than loose ad-hoc partnerships.
Research keywords
Highlights
PSMP-CLIP: patch-prompt SAM with multi-semantic prompting for zero-shot anomaly detection
TCNeRV: dual-domain temporal context modeling for implicit neural video compression
Timeline
- 2006The earliest works in this name-resolved OpenAlex record appear, with 2 works recorded in 2006.
- 2012Early-period output remains sparse, closing with 1 recorded work in 2012 after intermittent counts from 2008 to 2011.
- 2022Recorded activity resumes with 1 work in 2022 after a multi-year gap in the counts-by-year series.
- 2024Output rises to 2 recorded works in 2024, a level matched again in 2025.
- 2026Most active year on record with 3 works; two arXiv papers — PSMP-CLIP on zero-shot anomaly detection and TCNeRV on neural video compression — enter the Evolve catalog, both with Zhang as last author.
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 figures are drawn from a single OpenAlex snapshot; consolidated citation totals and output percentiles are unavailable, so the output and impact axes are not scored.
- "Shanjun Zhang" is a common name: this profile is name-resolved from public bibliographic data, and topic clusters such as "Asymmetric Synthesis and Catalysis" may reflect works by different researchers sharing the name.
- Only the two 2026 arXiv papers have been reviewed by Evolve; other works in the OpenAlex record are unassessed and venue information is absent.
- Neither cataloged paper releases code, so reproducibility rests on paper-level reporting (mean readiness 55 across 2 papers).