Explainable AI for Biodiversity Monitoring and Ecological Image Analysis
面向生物多样性监测与生态图像分析的可解释人工智能
机构 * Pratt School of Engineering, Duke University(杜克大学工程学院) ; British Antarctic Survey(英国南极调查局) ; Department of Applied Mathematics and Theoretical Physics, University of Cambridge(剑桥大学应用数学与理论物理系) ; Duke University Division of Marine Science and Conservation, Nicholas School of the Environment, Duke University Marine Laboratory(杜克大学海洋科学与保护分会,尼古拉斯环境学院,杜克大学海洋实验室)
AI总结 提出将可解释人工智能(XAI)作为生态模型验证的标准组件,通过案例展示其在图像分类、目标检测和分割中识别生物信号、揭示伪影、指导数据策略的能力,提升AI生态证据的可靠性。