Security Enhancement using Machine Learning for Concealed Object Detection

此EngagePlus是Engage Grant启动的研究项目的继续，该项目涉及卡尔加里大学的生物识别技术实验室(BTLab)和爱国者一号技术公司(PatriotOne)。该项目涉及通过开发非法隐藏物体检测系统来加强公共场所的安全。BTLab的研究人员将使用RGB深度传感器收集的步态和身体动作等软生物特征应用于辅助爱国者。其中一个正在开发使用超宽带认知雷达的对象检测。建议的Engage Plus Grant的两个主要目标如下：(1)将开发的算法和使用RGB深度传感器收集的深度数据与Partriot One实验数据收集协议完全集成。(2)继续研究和开发生成性对抗网络，以帮助减少训练PatrioOne系统所需的数据收集的繁琐任务BTLab和PatriotOne在Engage Plus期间的合作将导致研究和开发一种新的工艺，以设计和测试实时拼接目标探测的雷达系统。新的设计工艺将帮助PatriotOne改进他们的系统，带来最终的商业产品。

This EngagePlus is a continuation of the research project started in Engage Grant, involving the BiometricTechnologies Laboratory (BTLab) at the University of Calgary and Patriot One Technologies Inc. (PatriotOne).The project is related to enhancing security in public places by developing systems for illicit concealed objectsdetection. The BTLab researchers applied soft biometrics such as gait and body action, collected usingRGB-depth sensors, to assist PatriotOne is developing their object detection using an ultra-wide band cognitiveradar.The two main objectives of the proposed Engage Plus Grant are as follows:(1) Fully integrate the developed algorithms and the depth data collected using RGB-depth sensors with thePartriotOne experimental data collection protocol.(2) Continue research and development on Generative Adversarial Networks for helping reduce thecumbersome task of data collection for training the PatrioOne system.Ultimately, the collaboration between BTLab and PatriotOne during the Engage Plus will result in the researchand development of a novel process to design and test radar systems for concelaed object detection in real-time.The novel design process will assist PatriotOne in improving their system to bring to a finalized commercialproduct.