Archive for the 'TED' Category TED视频:Tan Le 能读懂脑电波的耳机

 

 

【内容概要】:

Tan Le 的发明主要是试着去理解我们脑部所发出的电波——我们的控制和体验中心。
当然,他们遇到了许多问题。其中最重要的两个如下:
1检测算法:脑神经复杂,大脑重要功能分散,大脑表面的褶皱也影响信号收集。
解决方法:展开皮层,接近更深的地方——源头。
2.将电极布在头上,要电胶对头皮处理,用户感受不好,而且比较贵
解决方法:装置是一个14声道保真的电脑采集系统,不需要对头皮进行处理,信号稳定,可以自由移动,且只要几百美元。检测算法:面部表情是意想不到的,通过敏感的表情调整,可以使之个性化。她介绍了一套认知系统:让你用意念来移动物体。电脑会在不同的想法中区分不同的差别。
她还说一旦你训练好了探测功能,这个功能可以背分享到任何一种电脑和应用平台。。
比如在游戏虚拟人物受意念控制。颜色灯光声音特效。只能家庭,开关窗帘、开关灯。轮椅根据表情移动,前后左右。
你也可以加入更多的探测方式

 

【评论】:

Erik Granger

Aug 26 2010: This is a terrible idea. People can't control their thoughts, how do you expect them to control a computer with their thoughts? Also, there was obviously concious intention when he was making the box fade. The entire purpose of this thing was to elliminate concious intention, she says so at the beginning of the video. Awful, terrible, horrible, evil idea. Will not buy from seller again.

这是一个可怕的想法,人们不能控制他们的想法,你怎么能够期待他们通过他们的意念去控制一个电脑?还有,在他让盒子消失的时候一定有潜意识。对于这个东西的整个是建立在忽略了他的潜意识的基础之上的,她在视频的开始就这样说。一个可怕的想法。不想再去买了。

  • Nils Dock

    Sep 15 2010: she explains that there are diferent types of uses in the program she used to test the cube she chose the setting for controling motion and moving things. the diversity of thoughts is just a reflection of our mind and while training it to understand your mind, you can lern to controll it. buddist monks seek a stillness of mind to seek clairvoiance. there is nice potential while using this device, when you sleep have it type your dreams.

  • 她解释了在她的项目里哪里有不同的种的用户,她曾经用他们来测试立方体。她测试了控制情感和移动物体的设置。在我们训练它去懂得你的想法时想法的差异是对我们想法的一个映射,你可以学习去控制它。蒙克斯兄弟寻求一个静止的想法去寻求清楚的脑电波明码(clairvoiance)。当使用这个设备时有非常好的潜力,当你睡觉时可以用它来记录你的梦。

    • Bobby Chauksay

      Sep 21 2010: I like the idea of typing one's dreams. To what I think, dreams can tell a great deal about what goes on in someone's sub-conscious and let them figure out the problems in their lives. This device definitely has great potential in gaming, helping the disabled and so on

    • 我很喜欢这个点子“记录一个人的梦”。我认为,梦可以解释很多关于在一个人的潜意识里发生着什么,并且帮助他们解决他们生活中存在的问题。这个设备肯定会有很大的潜力去帮助赌博、帮助残障人士等等。

    • Vilgot Huhn 10+

      Sep 21 2010: I'm no neurologist but it seems you have to be really clear and focused on a single thought for it to be able to read it. And since everyones brain looks a bit different (kind of like a finger print) you would have to teach the headset/program all of your thoughts so that it can recognize it. I don't know but I think learning the headset all your possible thoughts would be hard and take a lot of time. Furthermore I'm not sure if the thoughts you have when you dream are so "focused" that the headset to understand.
      I'm just speculating but reading our dreams doesn't seem probable to me.

    • 我不是一个神经病学家,但是看来你必须真正清楚地和专注地研究一个单独的想法并能够去读取它。但是每个人的大脑看起来肯定会有一些不同(就像指纹一样)你需要去教这个设备所有你的想法,以让他能够认出它。虽然我不懂,但是我认为学习所有你可能的想法回事非常难的和花时间的。此外,我不确定如果当你做梦的时候拥有的想法这个设备能读懂。我只是深思和推测,但是读取我们的梦对我来说是不可能的。

      • lo napala

        Nov 15 2010: We know that specific areas of the brain are activated when we feel love or fear and when we move our hands or walk. i wonder as this technology advances that all or most normal control subjects activate the same parts of the brain when thinking simple commands. Therefore the device would need little or no calibration. Possibly from there more difficult commands could be customized?

    • Mary Loescher
    • 1 day ago: But there are many similarities in how people think actions and images. And I'm sure pattern recognition can do nice things, as well as crowd sourcing, data mining, machine learning, etc. So eventually I imagine these things may be possible. This is a first baby step towards... who knows...

    • 但是在人们想到的动作和图片里有很多相同的东西。我确定模式识别可以做很棒的事。就像拥挤纯源化,数据挖掘,机器学习等等。所以最终,我想像这些东西都是可能的。谁会知道这是不是一个伟大的第一步呢?

 

【个人感觉】

个人感觉吧,这又是一个有意思的计算机应用,有点模式识别技术的感觉。先建立一下每个人的意识的模型,也就是让机器获取一下先验知识。然后利用这些已有的数据(模式)去识别它。我想,如果建立了足够多的模型,是不是就可以进一步虚拟出人们想像的世界啦,或者可以用它来帮助作家、科学家什么的记录下瞬间脑间所想,就像第二位评论者所说的记录梦境一样。的确蛮有新意的。

至于第一位评论者所说的人不能控制想法更不能控制电脑,这也未必,你只要建立足够多的模型,计算机就可以了解你的想法了,况且,这些研究可能只是启蒙,后续或许随着其它技术比如人工智能啊什么的的进步,也会促进此项技术的进步。

posted @ 2012-01-20 18:25  Ю詺菛╀時代  阅读(322)  评论(0编辑  收藏  举报