CAREER: Audio RESCUE: Realistic Audio for RESponders in Complex Undetermined Environments

职业：音频救援：在复杂的不确定环境中为响应者提供逼真的音频

基本信息

批准号：
1845324
负责人：
Kyla McMullen
金额：
$ 55万
依托单位：
University of Florida
依托单位国家：
美国
项目类别：
Continuing Grant
财政年份：
2019
资助国家：
美国
起止时间：
2019-08-15 至 2025-07-31
项目状态：
未结题

来源：
https://www.nsf.gov/awardsearch/showAward?AWD_ID=1845324&HistoricalAwards=false
关键词：
CAREER Audio RESCUE Realistic RESponders

项目摘要

First responders save lives in hazardous conditions such as hurricanes, tsunamis, earthquakes, fires, and terror events, and their role is more important than ever. For firefighters, the greatest challenges in search and rescue tasks are low visibility, disorientation, and tracking other team members. Despite methods to overcome these challenges, firefighters are still plagued by getting lost in buildings, miscommunication of locations and spaces, and not being able to identify paths. This project investigates using spatial (or 3D) sound to represent targets of interest (for example, a lost child) in order to allow first responders to use sound to navigate to specific points that they are unable to see. Using 3D sound in this context is a promising solution because responders' hands are often occupied, and a visual display can increase cognitive load. In addition, the present work will develop supplemental augmented-reality modules for firefighter emergency medical service (EMS) training. This training will allow firefighters to practice EMS skills in a realistic environment outside of the classroom. The PI and team will engage a diverse pool of underrepresented minority students through recruitment from the Distributed Research Experiences for Undergraduates (DREU) program and the Institute for African American Mentoring in Computing Sciences (iAAMCS). University of Florida (UF) undergraduate students will also be recruited through the Emerging Scholars Program to participate in the proposed work. Furthermore, firefighters will receive more interactive educational training as a direct result of their collaboration in this research program.This research will pursue a novel research program to create a fundamental understanding of how to move 3D audio research out of the laboratory into the real world. Working closely with Gainesville Fire Rescue (GFR) as the testbed application domain, the project team will evaluate using 3D audio to support a search and rescue mission. The PI and team will evaluate firefighters' performance and use the results to inform data-driven and empirically validated 3D sound design guidelines that address the following challenges: Realistic 3D sound rendering - this is challenging due to the individualistic nature of 3D sound filters, sound delivery hardware, and perceptual qualities of sounds; Change detection - in real-world contexts, a user must be able to quickly detect 3D sound changes, however, limited literature exists in this area; Effects of competing sounds - real-world scenarios are filled with background noise that may hinder the listener from accurately distinguishing a 3D sound of interest. Project outcomes will help first responders unobtrusively perceive the 360 degree world around them and maintain awareness of their surroundings.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

急救人员挽救了危险条件，例如飓风，海啸，地震，火灾和恐怖事件，其作用比以往任何时候都重要。对于消防员来说，搜索和救援任务中最大的挑战是知名度低，迷失方向和跟踪其他团队成员。尽管方法克服了这些挑战，但消防员仍然因在建筑物中迷失，位置和空间的沟通不畅而困扰，并且无法识别道路。该项目使用空间（或3D）声音进行调查，以表示感兴趣的目标（例如，失去的孩子），以允许急救人员使用声音导航到他们看不见的特定点。在这种情况下，使用3D声音是一个有前途的解决方案，因为响应者的手经常被占据，并且视觉显示可以增加认知负荷。此外，目前的工作将开发用于消防员紧急医疗服务（EMS）培训的补充增强现实模块。这项培训将使消防员可以在课堂外的现实环境中练习EMS技能。 PI和团队将通过从分布式研究经验（DREU）计划和非裔美国人指导的计算科学研究所（IAAMC）的分布式研究经验中招募各种代表性不足的少数族裔学生。佛罗里达大学（UF）本科生也将通过新兴学者计划招募，以参加拟议的工作。此外，消防员将接受更多的互动教育培训，这是他们在该研究计划中的合作的直接结果。这项研究将探讨一项新颖的研究计划，以对如何将3D音频研究从实验室移出到现实世界中有基本的了解。项目团队将与Gainesville Fire Rescue（GFR）紧密合作，作为测试台应用程序域，将使用3D音频评估搜索和救援任务。 PI和团队将评估消防员的性能，并使用结果为数据驱动和经验验证的3D声音设计指南提供信息，以应对以下挑战：现实3D声音渲染 - 由于3D声音过滤器的个人主义性质，声音，这是充满挑战的交付硬件和声音的感知品质；更改检测 - 在实际情况下，用户必须能够快速检测3D声音变化，但是，该领域存在有限的文献。竞争声音的效果 - 现实世界中的场景充满了背景噪音，这可能会阻碍听众准确区分3D感兴趣的声音。项目成果将帮助急救人员毫不客气地感知周围的360度世界并保持对周围环境的认识。该奖项反映了NSF的法定任务，并被认为是值得通过基金会的知识分子和更广泛影响的评估评估标准的评估值得支持的。

项目成果

期刊论文数量（3）

专著数量（0）

科研奖励数量（0）

会议论文数量（0）

专利数量（0）

Prediction of Object Geometry from Acoustic Scattering Using Convolutional Neural Networks

DOI：
10.1109/icassp39728.2021.9414743
发表时间：
2020-10
期刊：
ICASSP 2021 - 2021 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)
影响因子：
0
作者：
Ziqi Fan;Vibhav Vineet;Chenshen Lu;K. McMullen
通讯作者：
Ziqi Fan;Vibhav Vineet;Chenshen Lu;K. McMullen

Computation of Head-Related Transfer Functions Using Graphics Processing Units and a Perceptual Validation of the Computed HRTFs against Measured HRTFs

使用图形处理单元计算与头部相关的传递函数，并对计算的 HRTF 与测量的 HRTF 进行感知验证