Thermal Facial Analysis for Deception Detection
Thermal Facial Analysis for Deception Detection
复制标题
DOI：
10.1109/tifs.2014.2317309
复制
发表时间：
2014-06-01
影响因子：
6.8
通讯作者：
Zwiggelaar, Reyer
中科院分区：
文献类型：
--
作者：
Rajoub, Bashar A.;Zwiggelaar, Reyer
热成像技术可用于根据面部辐射的热量来检测人体的压力水平。在本文中，我们使用热成像来监测眶周区域的热变化，并测试它是否可以提供用于检测欺骗的判别特征。我们首先概述了自动欺骗检测，并提出了一种新颖的方法，我们对从 25 名参与者中提取的 492 个热响应（249 个谎言和 243 个真相）进行了实验验证。本文的新颖之处在于对每个主题的大量问题进行评分，强调从数据中学习的内部方法，提出验证决策过程的框架以及正确评估泛化性能。 k 最近邻分类器用于使用不同的数据表示策略对热响应进行分类。我们报告称，基于内部方法和五重交叉验证，预测谎言/真相反应的能力为 87%。我们的结果还表明，用于建模欺骗的人与人之间的方法并不能很好地概括整个训练数据。
Thermal imaging technology can be used to detect stress levels in humans based on the radiated heat from their face. In this paper, we use thermal imaging to monitor the periorbital region's thermal variations and test whether it can offer a discriminative signature for detecting deception. We start by presenting an overview on automated deception detection and propose a novel methodology, which we validate experimentally on 492 thermal responses (249 lies and 243 truths) extracted from 25 participants. The novelty of this paper lies in scoring a larger number of questions per subject, emphasizing a within-person approach for learning from data, proposing a framework for validating the decision making process, and correct evaluation of the generalization performance. A k-nearest neighbor classifier was used to classify the thermal responses using different strategies for data representation. We report an 87% ability to predict the lie/truth responses based on a within-person methodology and fivefold cross validation. Our results also show that the between-person approach for modeling deception does not generalize very well across the training data.