I-Corps: Real-Time Facial Tracking for Safety, Security and Medical applications

I-Corps：用于安全、安保和医疗应用的实时面部跟踪

• 批准号: 1355304
• 负责人: Dimitris Metaxas
• 金额: $ 5万
• 依托单位: Rutgers University New Brunswick
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2014
• 资助国家: 美国
• 起止时间: 2014-06-01 至 2015-05-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=1355304&HistoricalAwards=false
• 关键词: Corps; Real; Time; Facial; Tracking

Facial expressions play a significant role in our daily communication. However, few accurate, robust and portable expression detection and analysis systems currently exist. The project team has developed a system for real-time tracking and analyzing of the facial features of humans in video. The system, composed of software and a single camera can automatically analyze facial expressions, detect eye blinks and track the head orientation of a human target under varying conditions of illumination. With facial expression analysis, the barrier of machine detection of human emotion, which has existed since the invention of the earliest machines, could finally be lifted. This possibility allows the research team to explore not only the social innovations, such as expression based rating of media or sentiment of a social interaction, but the connection of human and machine through non-verbal, indirect communication. The recognition of human facial expressions through an automated system has extensive applications, such as; emotional reaction measurement, development of a facially controlled human-computer interface and a multitude of opportunities by releasing an individually configured input mapped facially controlled software interface to the public. Some of the possibilities of facial expression analysis could include enhanced search results based on the users approval, automatic word definition triggered by a "confused" expression and general and/or specific sentiment after the review of textual or video content. Through the fatigue detection capabilities of our software, the transportation industry can be made safer by the development of systems that can detect the alertness of pilots, drivers and vehicle operators at large and pair them with devices to alert them to their fatigued state.

面部表情在我们的日常交流中扮演着重要的角色。然而，目前几乎没有精确、稳健和便携式的表达检测和分析系统。 该项目团队开发了一种用于实时跟踪和分析视频中人类面部特征的系统。该系统由软件和单个摄像头组成，可以自动分析面部表情，检测眨眼，并在不同的照明条件下跟踪人类目标的头部方向。通过面部表情分析，机器检测人类情感的障碍，自最早的机器发明以来一直存在，最终可以被解除。这种可能性使研究团队不仅可以探索社会创新，例如基于表达的媒体评级或社交互动的情感，还可以通过非语言的间接沟通来建立人与机器的联系。通过自动化系统对人类面部表情的识别具有广泛的应用，例如：情绪反应测量、面部控制的人机接口的开发以及通过向公众发布单独配置的输入映射面部控制的软件接口的大量机会。面部表情分析的一些可能性可以包括基于用户批准的增强的搜索结果、由“困惑”表情触发的自动词定义以及在文本或视频内容的审查之后的一般和/或特定情绪。通过我们软件的疲劳检测功能，可以通过开发可以检测飞行员，驾驶员和车辆操作员的警觉性并将其与设备配对以提醒他们疲劳状态的系统，使运输行业更加安全。