Spotted: Location-informed Reidentification of Hyenas and Leopards in Camera Trap Surveys
作者: Halil Sina Kelebek, Julia Hindel, Kobus Hoffman, Lauren Hoffman, Andrew Loveridge, Bob Mandinyenya, Kudakwashe Ncube, Justin Seymour-Smith, Andrea Sibanda, Abhinav Valada, Matthew Wijers, Daniele De Martini
分类: cs.CV
发布日期: 2026-07-01
💡 一句话要点
提出Spotted框架以解决相机陷阱动物重识别问题
🎯 匹配领域: 支柱九：具身大模型 (Embodied Foundation Models)
关键词: 动物重识别 相机陷阱 时空可行性 视觉相似性 生态监测 深度学习 人机协作
📋 核心要点
- 现有动物重识别方法在相机陷阱调查中面临低图像质量和个体观察不平衡等挑战，导致性能不足。
- 本文提出Spotted框架，结合视觉相似性与时空可行性先验，减少专家审核的工作量。
- 在三个数据集上，Spotted模型的平均前五名识别准确率分别提高了9pp、2pp和9pp，且减少了69pp的查询比较次数。
📝 摘要（中文）
动物重识别（ReID）在相机陷阱调查中面临低图像质量、光照和视角变化大以及个体观察数量不平衡等挑战，导致现有ReID性能不足以实现完全自动化。本文提出Spotted框架，结合视觉相似性与来自相机位置的时空可行性先验，减少专家审核需求。该方法计算基于最小旅行速度的可行性评分，利用这些评分进行轻量级模型训练，并融合视觉相似性与时空可行性以获得稳健的配对匹配评分。我们在三个挑战性数据集上评估Spotted，显著提高了识别准确率。
🔬 方法详解
问题定义：本文旨在解决相机陷阱动物重识别中的低图像质量、光照变化和个体观察数量不平衡等问题。现有方法主要依赖视觉线索，忽视了可用的时空信息，导致自动化程度低。
核心思路：Spotted框架通过结合视觉相似性与时空可行性先验，利用相机位置和时间戳信息，减少了对专家审核的依赖。该设计旨在提高重识别的准确性和效率。
技术框架：整体架构包括三个主要模块：首先，计算基于最小旅行速度的可行性评分；其次，利用这些评分进行轻量级模型的训练；最后，融合视觉相似性与时空可行性以获得配对匹配评分。
关键创新：最重要的创新点在于引入时空可行性先验作为伪监督，结合视觉信息进行训练，显著提高了重识别的准确性和效率。与现有方法相比，Spotted框架更全面地利用了可用信息。
关键设计：在模型设计中，采用了轻量级网络结构以适应实时应用，同时设置了合适的损失函数来平衡视觉相似性与时空可行性评分的融合。
🖼️ 关键图片
📊 实验亮点
在三个挑战性相机陷阱数据集上，Spotted模型的平均前五名识别准确率分别提高了9pp、2pp和9pp，显著优于最佳基线。同时，采用人机协作策略使得查询比较次数减少了69pp，提升了整体工作效率。
🎯 应用场景
该研究的潜在应用领域包括野生动物监测、生态研究和保护生物学等。通过提高动物重识别的准确性和效率，Spotted框架能够帮助研究人员更好地理解动物行为和生态动态，进而为保护工作提供数据支持。
📄 摘要（原文）
Animal re-identification (ReID) in camera-trap surveys remains challenging due to low image quality, strong variation in illumination and viewpoint, and highly imbalanced numbers of observations per individual. As a result, current ReID performance is often insufficient for fully automated use, and practical workflows typically depend on expert review of algorithmically proposed candidate matches. Moreover, most existing approaches focus almost exclusively on visual cues and overlook auxiliary information routinely available in field studies, such as image timestamps and camera-trap locations. We introduce Spotted, a location-informed, human-in-the-loop animal ReID framework that integrates visual similarity with spatio-temporal feasibility priors derived from camera locations, thereby reducing the amount of required expert review. Our method (i) computes an image-model-agnostic feasibility score based on the minimum travel speed required for two detections to correspond to the same individual, (ii) uses these feasibility cues as pseudo-supervision to train a lightweight head on top of a frozen visual foundation model, and (iii) fuses adapted visual similarity with spatio-temporal feasibility to obtain a robust pairwise matching score. We additionally integrate an active pair sampling strategy to accelerate annotation by initially prioritizing uncertain predictions. We evaluate Spotted on three challenging camera-trap ReID datasets comprised of spotted hyenas and leopards, which we release as part of this work. Our model improves average top-5 identification accuracy by 9pp, 2pp and 9pp over the best baseline on our LeopardID102, SpottedHyenaID109 and SpottedHyenaID415 datasets, respectively. Further, we show that our human-in-the-loop strategy reduces the number of queried comparisons by up to 69pp while achieving equivalent positive matches.