(单词翻译:单击)
听力文本
Scientists may soon be able to interpret what someone is saying simply by analysing their brainwaves as they speak. This revolutionary advance in neuroscience would help millions of people who suffer from communication problems and neurological disorders. The scientists developed a form of artificial intelligence that can decode brainwaves and translate them into text. Algorithms take the brain activity created as a person speaks and translates it in real time into sentences on a screen. The scientists are from the University of California, San Francisco. They say their algorithms have a 97 per cent translation accuracy rate but are working hard to improve on this.
The scientists say they are at the early stages of being able to machine-translate everything someone says. The software used in their experiments matched features of speech that were repeated frequently to parts and shapes of the mouth. These included elements of English speech such as vowels, consonants and commands. The experiments were limited to around 40 short and simply-constructed spoken sentences. The scientists said: "Although we should like the decoder to learn and exploit the regularities of the language, it remains to show how many data would be required to expand from our tiny languages to a more general form of English."
译文由可可原创，仅供学习交流使用，未经许可请勿转载。
重点解析
1.Scientists may soon be able to interpret what someone is saying simply by analysing their brainwaves as they speak.
interpret 解释；阐明
The whole speech might well be interpreted as a coded message to the Americans
整个演讲完全可以看成是在向美国人传递编码信息。
2.The scientists developed a form of artificial intelligence that can decode brainwaves and translate them into text.
decode 解码
All he had to do was decode it and pass it over
他需要做的就是将它破译然后转给他人。
参考翻译
科学家们可能很快就能通过分析说话人的脑电波来翻译他们所说的话。神经科学的这一革命性进展将帮助数百万患有沟通障碍和神经障碍的人。科学家们开发了一种人工智能，能够解码脑电波，将其转换成文字。这些算法将人说话时产生的大脑活动实时转换成屏幕上的文字。这群科学家来自加州大学旧金山分校。他们表示他们算法的翻译准确率为97%，但他们正在努力改进。
这些科学家们表示，要让机器能够翻译人们的每句话，他们还处在该目标的早期阶段。他们实验中所用的软件与嘴部和嘴巴形状频繁重复的语言特征相匹配。这些包括英语演讲的元素，如元音、辅音和祈使句。实验限定在约40个简短且结构简单的的口语句子。科学家们表示：“尽管我们希望解码器能学习和利用语言的规律，但仍需证明需要多少数据才能从我们的极少量的语言扩展到更常见的英语形式。”
