---
格式版本: 2
标题: "SPARC: Subspace Position-Aware Robust Few-Shot Calibration for Distribution-Shifted Industrial Anomaly Detection"
原文链接: "https://arxiv.org/abs/2608.18585"
发布日期: "2026-08-19"
发布时间校准状态: "found"
发布时间需复核: "否"
发布时间来源: "rule:local:strict_original_body"
发布时间证据: "**\\[v1\\]** Wed, 19 Aug 2026 06:25:35 UTC (14,672 KB)"
发布时间校准原因: "规则确认唯一严格发布时间，来源 local:strict_original_body"
发布时间校准置信度: "high"
发布时间候选数量: 18
发布时间严格候选数量: 6
发布时间原页读取状态: "source template page reused from URL open"
发布时间未找到原因: ""
发布时间校准时间: "2026-08-20T15:23:47+08:00"
发布时间仲裁状态: "skipped"
发布时间仲裁尝试次数: 0
发布时间仲裁耗时毫秒: 0
发现时间: "2026-08-20T15:18:01+08:00"
入库时间: "2026-08-20T07:23:47.788Z"
来源平台: "arXiv 学术论文搜索"
搜索渠道: "source_template"
搜索词: "https://arxiv.org/search/?query=deployment&searchtype=all"
匹配关键词:
  - "deployment"
相关厂家:
  []
相关专家:
  []
内容类型: "网页"
抓取工具: "Free Fetch + Defuddle"
清洗工具: "Defuddle Markdown + Defuddle/Readability 正文提取"
原始附件:
  []
AI优质: "否"
AI打分: 7
AI分档: "非优质"
AI质检状态: "不通过"
AI打分理由: "该资料为arXiv论文页，主题是工业异常检测的少样本校准，与超节点/AI Rack/机柜级AI基础设施完全无关。仅搜索词‘deployment’命中，属于明显无关内容。"
AI质检模型: "ali-deepseek-v4-flash"
AI质检时间: "2026-08-20T15:28:24+08:00"
AI主题相关性: 0
AI来源权威性: 5
AI新颖性: 0
AI技术细节: 0
AI商业部署信号: 0
AI完整性: 2
AI摘要: "SPARC提出一种面向分布偏移的工业异常检测少样本校准方法，在编码器与检测器之间拦截补丁特征，通过逐单元子空间投影去除部署时的干扰，无需梯度更新，仅需少量已验证正常图像即可适配。"
AI摘要模型: "ali-deepseek-v4-flash"
AI摘要时间: "2026-09-07T03:18:34.278Z"
采集批次: "2026年8月20日14点19分32秒"
采集批次ID: "20260820-141932-079"
去重键: "https://arxiv.org/abs/2608.18585"
---

## Computer Science > Computer Vision and Pattern Recognition

## Title:SPARC: Subspace Position-Aware Robust Few-Shot Calibration for Distribution-Shifted Industrial Anomaly Detection

Authors:[Seokhee Han](https://arxiv.org/search/cs?searchtype=author&query=Han,+S), [Seungjun Chu](https://arxiv.org/search/cs?searchtype=author&query=Chu,+S), [Mateusz Nowak](https://arxiv.org/search/cs?searchtype=author&query=Nowak,+M), [Peter Chin](https://arxiv.org/search/cs?searchtype=author&query=Chin,+P)

[View PDF](https://arxiv.org/pdf/2608.18585) [HTML (experimental)](https://arxiv.org/html/2608.18585v1)

> Abstract:Vision-based industrial anomaly detectors are calibrated on one distribution but may be deployed on another that differs in illumination, fixture placement, or sensor characteristics, sharply degrading an otherwise accurate detector. Adapting to the incoming lot is a natural response, but labeled anomalies are scarce. We therefore consider calibration using only a handful of verified-normal images available before scoring the rest of the lot. Existing fixes require backpropagation, detector-specific tuning, or choices about feature directions that few calibration samples cannot justify. We present SPARC, a few-shot calibration method that intercepts patch features between encoder and detector and removes a closed-form, spatially indexed estimate of deployment-time nuisance through per-cell subspace projection. It needs only $k \le 8$ verified-normal images and uses the algebraic saturation rank $r{=}k{-}1$ on the encoder's native patch grid. The correction requires no gradient or weight updates and works with memory-bank, density, prototype, and mutual detectors. On the shift-prone benchmarks, SPARC improves pooled Image AUROC and AU-PRO $_{0.3}$ for all seven detectors whose image scores depend on corrected patch features by $+13.8$ and $+3.5$ percentage points (pp), respectively; on benchmarks without engineered shift, the changes are small and mixed. Controls that give competing corrections the same calibration images attribute these gains to the per-cell subspace structure rather than the images alone. Further ablations support the saturation-rank choice and characterize sensitivity to backbone and calibration conditions.

| Subjects: | Computer Vision and Pattern Recognition (cs.CV) |
| --- | --- |
| Cite as: | [arXiv:2608.18585](https://arxiv.org/abs/2608.18585) \[cs.CV\] |
|  | (or [arXiv:2608.18585v1](https://arxiv.org/abs/2608.18585v1) \[cs.CV\] for this version) |
|  | [https://doi.org/10.48550/arXiv.2608.18585](https://doi.org/10.48550/arXiv.2608.18585) |

## Submission history

From: Seokhee Han \[[view email](https://arxiv.org/show-email/f1d2c4f5/2608.18585)\]  
**\[v1\]** Wed, 19 Aug 2026 06:25:35 UTC (14,672 KB)

[Which authors of this paper are endorsers?](https://arxiv.org/auth/show-endorsers/2608.18585) | Disable MathJax ([What is MathJax?](https://info.arxiv.org/help/mathjax.html))