首页> 外文会议>Simulation Interoperability Workshop >Building an Angry Grandmother
【24h】

Building an Angry Grandmother

机译:建立一个愤怒的祖母

获取原文
获取外文期刊封面目录资料

摘要

Historically, the Infantry Immersion Trainer has relied heavily on human role-players to serve as the occupants of a simulated village located in Iraq or Afghanistan. While this works for many situations, it is expensive. Additionally, there are some vital roles and situations that aren 't easily portrayed with live role-players (such as those of children or the elderly). The facility currently includes support for the use of mixed reality characters, which are displayed on life-sized screens in some areas of the village. These characters have been designed primarily for use in kinetic "shoot or no-shoot" scenarios requiring relatively short interactions with the trainees. Under the Future Immersive Training Environment Joint Capability Technology Demonstration, one goal was to develop technology which could be used to create characters who can force the trainees to make a variety of complex decisions in non-kinetic situations, as part of scenarios that last for an extended period of time (10 minutes or more). These characters need to be visually and behaviorally plausible, culturally appropriate, and responsive to the actual situation and the actions of the trainees. They should be able to act autonomously, but also be responsive to operator control. Our test character portrays an elderly Afghan woman who is the grandmother of an insurgent. She is built around a modular artificial intelligence which works in conjunction with improved artwork, a more complex animation system, instrumentation of the room and the trainees, and direction from the Instructor/Operator Station. She can not only react appropriately to the actions of the Marines, but can also respond to questioning and engage in limited conversation (in her native language of Pashto), all while attempting to minimize the visible repetition of behavior. She has not only succeeded in forcing the trainees to make difficult decisions, but also leaves us well positioned to incorporate further improvements (such as more detailed instrumentation of the room, or a social/psychological model of the village state) in future characters.
机译:从历史上看,步兵浸入培训师依赖于人类角色玩家,作为位于伊拉克或阿富汗的模拟村庄的占用者。虽然这适用于许多情况,但它很昂贵。此外,还有一些重要的角色和情况,不容易与现场角色玩家(如儿童或老年人)的表现形式。该设施目前包括对使用混合现实人物的支持,这些字符显示在村庄的某些地区的寿命屏幕上。这些角色主要是用于在动力学“拍摄或无拍摄”情景中的设计,需要与学员相对较短的互动。在未来的沉浸式培训环境联合能力技术示范下,一个目标是开发技术,可以用来创造能够强迫受训人员在非动力学情况下制作各种复杂决策的人物,作为持续的情景的一部分延长时间(10分钟或更长时间)。这些人物需要在视觉上和行为地是合理的,文化合适的,并响应于实际情况和受训人员的行为。他们应该能够自主行动,而且对操作员控制响应。我们的测试角色描绘了一位老年阿富汗女人,他是一个叛乱的祖母。她围绕模块化人工智能建立,与改进的艺术品,更复杂的动画系统,房间仪器和学员以及来自教练/操作员站的方向相结合。她不仅可以适当地对海军陆战队的行为作出反应,而且还可以回应质疑和参与有限的对话(以她的汉语语言),同时尝试最小化行为的可见重复。她不仅成功地强迫受训人员做出困难的决定,而且还将我们的定位良好地纳入了未来的特征的进一步改进(例如房间的更详细的仪器或村庄的社会/心理模型)。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号