文本选自:The Economist(经济学人)
作者:Unknown
原文标题:Wi-Fi signals could prove useful for spies
原文发布时间:28 Jan. 2023
Wi-Fi signals could prove useful for spies
Like all radio waves, Wi-Fi signals undergo subtle shifts when they encounter objects—human beings included. These can reveal information about the shape and motion of what has been encountered, in a manner akin to the way a bat's chirps reveal obstacles and prey.
Starting from this premise Jiaqi Geng, Dong Huang and Fernando De la Torre, of Carnegie Mellon University, in Pittsburgh, wondered if they could use Wi-Fi to record the behaviour of people inside otherwise unobservable rooms. As they describe in a posting on arXiv, they have found that they can. "DensePose from Wi-Fi", the paper in question, describes how they ran Wi-Fi signals from a room with appropriate routers in it through an artificial-intelligence algorithm trained on signals from people engaging in various, known activities. This algorithm was able to reconstruct moving digital portraits, called pose estimations, of the individuals in the room.
Mr Geng, Dr Huang and Dr De la Torre are not the first to think of doing this. But they seem to have made a significant advance. Earlier experiments had managed to obtain two-dimensional (2d) pose estimations based on as many as 17 "vector points" on the body—such as head, chest, knees, elbows and hands. The new paper, by contrast, describes "2.5d" portraits that track 24 vector points. And, according Dr Huang, the team has now built an enhanced version capable of generating complete 3d body reconstructions that track thousands of vector points.
Wi-Fi信号可能被证明对间谍有用
精听党背景导读
“在某些情况下,Wi-Fi信号可以作为RGB图像的替代来进行人体感知。”
近日发布的论文《DensePose from Wi-Fi》中,三位来自CMU的研究者表达了这一结论。同时根据实验过程中所获得的相关数据,环境光线与信号遮挡对室内监控的Wi-Fi解决方案影响甚微。虽然在过去的几年间,由于自动驾驶和VR技术的广泛应用,通过2D或者3D传感器进行人体姿态估计取得较大进展,但传感器所具备的超高成本以及在日常情境下的较大用电需求都让人望而却步。
文本选自:The Economist(经济学人)作者:Unknown原文标题:Wi-Fi signals could prove useful for spies原文发布时间:28 Jan. 2023关键词:Wi-Fi 算法 人工智能
精听党带着问题听
1. 如何理解第二段中的“in question”?2.“取得重大进展”用英语如何表达?3. 根据本文,该团队的最新研究进展如何?
精听党选段赏析
标题解读
Wi-Fi signals could prove useful for spies
Wi-Fi信号可能被证明对间谍有用
段一
Like all radio waves, Wi-Fi signals undergo subtle shifts when they encounter objects—human beings included. These can reveal information about the shape and motion of what has been encountered, in a manner akin to the way a bat’s chirps reveal obstacles and prey.
undergo vt. 经历,经受;1. to undergo trials/repairs 进行试验/接受检修;
subtle adj. 不易察觉的;不明显的;微妙的;1. There are subtle differences between the two versions. 两个版本之间有一些细微的差异。
encounter vt. 遭遇;偶遇;
reveal vt. 揭示;显示;透露;1. to reveal a secret 泄露一条秘密;
akin to 类似;相当于;1. What he felt was more akin to pity than love. 他感受到的更像怜悯,而不是爱。
chirp n. 吱吱声,唧唧声;抱怨;故意惹怒(某人)而说,奚落;
prey n. 猎物,捕获物;1. predator n. 捕食性动物;掠夺者;
参考译文
与所有无线电波一样,Wi-Fi信号在遇到物体(包括人类)时会发生微妙的变化。它们可以揭示所遇到物体的形状和运动的信息,类似于蝙蝠的发声以揭示障碍物和猎物。
段二
Starting from this premise Jiaqi Geng, Dong Huang and Fernando De la Torre, of Carnegie Mellon University, in Pittsburgh, wondered if they could use Wi-Fi to record the behaviour of people inside otherwise unobservable rooms. As they describe in a posting on arXiv, they have found that they can. “DensePose from Wi-Fi”, the paper in question, describes how they ran Wi-Fi signals from a room with appropriate routers in it through an artificial-intelligence algorithm trained on signals from people engaging in various, known activities. This algorithm was able to reconstruct moving digital portraits, called pose estimations, of the individuals in the room.
premise n. 前提,假设;1. on the premise 在……前提下;
wonder vt. 想知道,好奇,琢磨;不知是否(用于提问或礼貌地提出请求);
otherwise adj. 别的;在其他各方面…的;在其他不同情况下的;
in question 讨论(或议论)中的;1. On the day in question we were in Cardiff. 在我们讨论的那一天,我们在加的夫。
artificial-intelligence algorithm 人工智能算法;
engage in 从事于(参加);1. Even in prison, he continued to engage in criminal activities. 即使在监狱里,他仍继续从事犯罪活动。
moving digital portrait 移动数字肖像;1. portrait n.(尤指只刻画脸、头和肩部的)肖像,照片,雕像;描绘,描述;
estimation n. 估计,估算;判断,评价;1. pose estimations 姿态估计;姿势判断;
参考译文
从这个前提出发,匹兹堡卡内基梅隆大学的耿佳琪、黄东和费尔南多·德·拉·托雷想知道他们是否可以用Wi-Fi来记录人们在其他无法观察的房间里的行为。正如他们在arXiv上的一篇文章中所描述的那样,他们发现他们可以做到。这篇论文名为《DensePose from Wi-Fi》,描述了他们如何通过人工智能算法,在一个有适当路由器的房间里运行Wi-Fi信号,该算法接受来自人们从事各种已知活动的信号的训练。这种算法能够重建房间内个体的移动数字肖像,称为姿势判断。
段三
Mr Geng, Dr Huang and Dr De la Torre are not the first to think of doing this. But they seem to have made a significant advance. Earlier experiments had managed to obtain two-dimensional (2d) pose estimations based on as many as 17 “vector points” on the body—such as head, chest, knees, elbows and hands. The new paper, by contrast, describes “2.5d” portraits that track 24 vector points. And, according Dr Huang, the team has now built an enhanced version capable of generating complete 3d body reconstructions that track thousands of vector points.
make a significant advance 取得重大进展;1. advance n. 前进;进步,进展;改善;
vector n. 向量,矢量;1. vector points向量点;矢量点;
chest n. 胸部,胸膛;
knee n. 膝盖,膝关节;(裤子的)膝部;1. knee joint 膝关节;铰链关节;
elbow n. 肘,肘部;
by contrast 相比之下;与之相比;
track vt. 追查,追踪;(尤指用特殊电子设备)跟踪,追踪;跟踪(进展情况);
capable of doing sth. 有能力做某事;
参考译文
耿先生、黄博士和德·拉·托雷博士并不是第一个想到这样做的人。但他们似乎已经取得了重大进展。早期的实验已经成功地获得了基于身体上多达17个“向量点”(如头部、胸部、膝盖、肘部和手部)的二维姿势判断。相比之下,这篇新论文描述了追踪24个向量点的“2.5维”肖像。根据黄博士的说法,该团队现在已经建立了一个增强版本,能够生成完整的三维身体重建,跟踪数千个向量点。
精听党每日单词
undergo
/ˌʌndərˈɡoʊ/ vt. 经历,经受;
subtle
/ˈsʌtl/ adj. 不易察觉的;不明显的;微妙的;
encounter
/ɪnˈkaʊntər/ vt. 遭遇;偶遇;
reveal
/rɪˈviːl/ vt. 揭示;显示;透露;
akin to
类似;相当于;
chirp
/tʃɜːrp/ n. 吱吱声,唧唧声;故意惹怒(某人)而说,奚落;
prey
/preɪ/ n. 猎物,捕获物;
premise
/ˈpremɪs/ n. 前提,假设;
wonder
/ˈwʌndər/ vt. 想知道,好奇,琢磨;不知是否(用于提问或礼貌地提出请求);
otherwise
/ˈʌðərwaɪz/ adj. 别的;在其他各方面…的;在其他不同情况下的;
in question
讨论(或议论)中的;
artificial-intelligence algorithm
人工智能算法;
engage in
从事于(参加);
moving digital portrait
移动数字肖像;
estimation
/ˌestɪˈmeɪʃn/ n. 估计,估算;判断,评价;
make a significant advance
取得重大进展;
vector
/ˈvektər/ n. 向量,矢量;
chest
/tʃest/ n. 胸部,胸膛;
knee
/niː/ n. 膝盖,膝关节;(裤子的)膝部;
elbow
/ˈelboʊ/ n. 肘,肘部;
by contrast
相比之下;与之相比;
track
/træk/ vt. 追查,追踪;(尤指用特殊电子设备)跟踪,追踪;跟踪(进展情况);
capable of doing sth.
有能力做某事;
精听党文化拓展
美国国家标准与技术研究所(NIST)的研究人员开发了一种基于Wi-Fi信号的微小变化来监测呼吸的方法。他们说,他们的BreatheSmart深度学习算法可以帮助检测家庭中是否有人有呼吸问题。
Wi-Fi信号几乎是无处不在的。它们在试图连接设备和路由器时,会从表面反弹并穿过。但是任何运动都会改变信号传递的路径,包括我们呼吸时身体的运动方式。如果我们有任何问题,这可能都会改变。例如,如果你在咳嗽,你的胸部会有不同的移动。
其他研究人员已经探索了如何使用Wi-Fi信号来检测人的运动,但他们的方法需要专用的传感设备,而且他们的研究提供的数据有限。几年前,一家名为Origin Wireless的公司开发了一种可与Wi-Fi网状网络配合使用的算法。同样地,NIST说BreatheSmart可以与市场上已有的路由器和设备一起使用。它只需要一个路由器和用于连接的设备。
科学家们改变了路由器上的固件,使其更频繁地检查 "频道状态信息",或CSI。CSI是指从一个设备,如电话或笔记本电脑,发送到路由器的信号。CSI信号是一致的,路由器了解它们应该是什么样子,但环境中的偏差,如信号受到表面或运动的影响,会修改信号。研究人员让路由器每秒请求这些CSI信号多达10次,以更好地了解信号是如何被修改的。
研究小组用一个人体模型模拟了几种呼吸条件,并用一个现成的路由器和接收设备监测CSI信号的变化。为了理解他们收集的数据,NIST的副研究员Susanna Mosleh开发了这个算法。在一篇论文中,研究人员指出,BreatheSmart在99.54%的时间里正确识别了人体模型所模拟的呼吸状况。
精听党每日美句
Better late than never.
迟做总比不做好。