KnifeHunter: Structured Local Representation Learning for Fine-Grained Knife Image Retrieval in Law Enforcement
作者: Syed Sameed Husain, Eng-Jon Ong, Stephen Simpson, Trevor Hamshere, Matt Turner, Miroslaw Bober
分类: cs.CV
发布日期: 2026-08-07
💡 一句话要点
提出KnifeHunter以解决执法领域刀具图像检索问题
🎯 匹配领域: 支柱二:RL算法与架构 (RL & Architecture)
关键词: 刀具图像检索 法医技术 深度学习 图像处理 机器学习
📋 核心要点
- 现有手动视觉比较方法在刀具识别中效率低下,难以扩展,且在操作成像条件下难以应用。
- 提出KnifeHunter系统及CoRe-Net架构,通过结构化互补表示学习和双向互融技术,提升刀具图像检索的准确性和效率。
- KnifeHunter在实际执法操作中表现优异,Medium协议下mAP达到88.0%,在现场查询中mP@1达到99.2%。
📝 摘要(中文)
刀具相关暴力行为对公共安全构成重大挑战,执法机构需要可扩展的工具进行刀具识别和情报分析。本文介绍了KnifeHunter,一个为英国执法机构开发的端到端法医刀具图像检索系统。该系统贡献了KnifeHunter数据集,包含25,843张来自警方证据、零售目录和边境查获的543个刀具类别的图像,并提供结构化元数据和评估协议。我们提出了CoRe-Net,一个紧凑的单描述符检索架构,结合了全局上下文与空间局部的判别证据。通过实验,KnifeHunter在实际应用中取得了99.2%的mP@1,展示了其在执法环境中进行精细法医刀具匹配的有效性。
🔬 方法详解
问题定义:本文旨在解决刀具图像检索中的效率和准确性问题,现有方法在实际执法环境中难以满足快速和高效的需求。
核心思路:KnifeHunter通过引入CoRe-Net架构,结合全局和局部特征,利用结构化互补表示学习来优化刀具图像的检索过程。
技术框架:KnifeHunter系统包括数据集构建、特征提取、检索模型训练和评估等模块。CoRe-Net作为核心模型,负责图像特征的提取与匹配。
关键创新:引入结构化互补表示学习(SCRL)和双向互融(BDRF)技术,能够有效整合全局和局部信息,显著提升检索性能。
关键设计:使用EVA02-Base作为骨干网络,结合余弦相似度进行检索,采用特定的损失函数和参数设置以优化模型性能。整体设计旨在提高检索的准确性和速度。
🖼️ 关键图片
📊 实验亮点
KnifeHunter在Medium协议下实现了88.0%的mAP和86.7%的mP@10,在干扰条件下也达到了85.1%的mAP和83.8%的mP@10。实际应用中,系统在现场查询中取得了99.2%的mP@1,展示了其卓越的检索性能。
🎯 应用场景
KnifeHunter系统具有广泛的应用潜力,特别是在执法机构的刀具暴力案件调查中。通过快速、准确的刀具识别,能够有效支持情报分析和证据管理,提升执法效率。未来,该系统可扩展至其他类型的物证检索,推动法医技术的发展。
📄 摘要(原文)
Knife-enabled violence presents a major public safety challenge, and law enforcement agencies require scalable tools for catalogue-level knife identification, intelligence analysis, and source attribution. Manual visual comparison is specialist, time-consuming, and difficult to scale under operational imaging conditions. We introduce KnifeHunter, an end-to-end forensic knife image retrieval system developed with UK law enforcement. The work contributes the KnifeHunter dataset, comprising 25,843 images across 543 knife classes from police evidence, retail catalogues, and border-force seizures, with structured metadata, Medium/Hard evaluation protocols, and large-scale distractor evaluation. We further propose CoRe-Net, a compact single-descriptor retrieval architecture that combines global context with spatially localised discriminative evidence. CoRe-Net introduces Structured Complementary Representation Learning (SCRL) to organise local evidence into complementary prototype-based representations, and Bi-Directional Reciprocal Fusion (BDRF) to integrate global and local evidence through residual projection and gated local-to-global injection. Using an EVA02-Base backbone and cosine-similarity retrieval, CoRe-Net achieves 88.0% mAP and 86.7% mP@10 on the Medium protocol, and 85.1% mAP and 83.8% mP@10 under distractor conditions. KnifeHunter was deployed by UK police forces during Operation Sceptre deployments from 2023 to 2025, achieving 99.2% mP@1 on field queries. These results demonstrate a practical and effective multimedia retrieval framework for fine-grained forensic knife matching in operational law-enforcement settings.