---
格式版本: 2
标题: "RailSyn: Diagnosis-Guided Image Generation for Traceable Data Completion in Railway Foreign Object Detection"
原文链接: "https://arxiv.org/abs/2608.30709"
发布日期: "2026-08-31"
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\\]** Mon, 31 Aug 2026 12:47:28 UTC (4,526 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: "2026-09-02T03:50:04+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: "2026-09-02T03:49:17+08:00"
入库时间: "2026-09-01T19:50:04.614Z"
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=Scale-up&searchtype=all"
匹配关键词:
  - "Scale-up"
  - "AI"
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: 5
AI分档: "非优质"
AI质检状态: "不通过"
AI打分理由: "该资料为arXiv计算机视觉论文，主题是铁路异物检测，与超节点/AI Rack/机柜级AI基础设施完全无关，仅标题命中Scale-up搜索词。"
AI质检模型: "zj-deepseek-v4-flash"
AI质检时间: "2026-09-02T03:50:29+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: 0
AI摘要: "RailSyn 提出一种诊断引导的图像生成框架，用于铁路异物检测的数据补全。"
AI摘要模型: "ali-deepseek-v4-flash"
AI摘要时间: "2026-09-01T23:06:26.956Z"
采集批次: "2026年9月1日23点56分52秒"
采集批次ID: "20260901-235652-761"
去重键: "https://arxiv.org/abs/2608.30709"
---

## Computer Science > Computer Vision and Pattern Recognition

## Title:RailSyn: Diagnosis-Guided Image Generation for Traceable Data Completion in Railway Foreign Object Detection

Authors:[Quan Hao](https://arxiv.org/search/cs?searchtype=author&query=Hao,+Q), [Chenxi Zhang](https://arxiv.org/search/cs?searchtype=author&query=Zhang,+C), [Ziyang Tao](https://arxiv.org/search/cs?searchtype=author&query=Tao,+Z), [Yuyuan Zhou](https://arxiv.org/search/cs?searchtype=author&query=Zhou,+Y), [Yudong Wang](https://arxiv.org/search/cs?searchtype=author&query=Wang,+Y), [Rui Shi](https://arxiv.org/search/cs?searchtype=author&query=Shi,+R), [Lechuan Xu](https://arxiv.org/search/cs?searchtype=author&query=Xu,+L), [Changhao Liu](https://arxiv.org/search/cs?searchtype=author&query=Liu,+C), [Liguo Zhang](https://arxiv.org/search/cs?searchtype=author&query=Zhang,+L)

[View PDF](https://arxiv.org/pdf/2608.30709) [HTML (experimental)](https://arxiv.org/html/2608.30709v1)

> Abstract:Railway foreign object detection (RFOD) is critical to safe railway operation, yet scarce real positive samples incompletely represent task-relevant variations in object scale, intrusion relation, railway scene, illumination, and adverse weather. Existing synthetic augmentation can improve RFOD detection, but its gains lack an explicit account of the task-relevant deficiencies complemented by the generated data. We therefore introduce RailSyn, a diagnosis-guided framework comprising a real-referenced Inspector and a requirement-aligned Generator. The Inspector constructs a variable-radius empirical cover from finite real observations to localize candidate completion regions and profile synthetic pools. The resulting audit identifies railway-context, intrusion-semantic, and visual-consistency requirements; the Generator addresses them through domain adaptation, agent-planned placement and physical contact relations, and plan-consistent conditional refinement. Using the Inspector, we further trace representation-space changes across generation variants; the complete system attains a local-shell occupation of $C_{gap}$ to 13.64%, which measures generated coverage of real-derived completion regions. Extensive experiments show AP50--95 gains of up to 4.9 points and consistent improvements across nine mainstream detectors, demonstrating broad cross-architecture utility.

| Subjects: | Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI) |
| --- | --- |
| Cite as: | [arXiv:2608.30709](https://arxiv.org/abs/2608.30709) \[cs.CV\] |
|  | (or [arXiv:2608.30709v1](https://arxiv.org/abs/2608.30709v1) \[cs.CV\] for this version) |
|  | [https://doi.org/10.48550/arXiv.2608.30709](https://doi.org/10.48550/arXiv.2608.30709) |

## Submission history

From: Ziyang Tao \[[view email](https://arxiv.org/show-email/e544c63e/2608.30709)\]  
**\[v1\]** Mon, 31 Aug 2026 12:47:28 UTC (4,526 KB)

[Which authors of this paper are endorsers?](https://arxiv.org/auth/show-endorsers/2608.30709) | Disable MathJax ([What is MathJax?](https://info.arxiv.org/help/mathjax.html))