评论
打印
After National Disabled Person's Day on May 21, a video released by CCTV featuring two graduate students from Beihang University and Tsinghua University has gone viral online.
在5月21日“全国助残日”之后，中央电视台发布的一段有关两位来自北航、清华的毕业生的视频，迅速在网上流传起来。
Wang Nana and Huang Shuang, intent on helping their hearing-impaired friends, invented a wearable sign-language interpreter by transforming sign language into audible information.
王娜娜和黄爽为了帮助自己听力受损的朋友，研究发明了一种能将手语转换为语音的翻译臂环。
The pair won first prize last month in an open design challenge organized by the UN Development Program and tech giant Baidu.
她们还于上个月在联合国开发计划署和科技巨头百度公司共同举办的设计挑战大赛上夺得了冠军。
Wang Nana started her research with a goal of helping the disabled rather than making a fortune.
王娜娜开始这一研究的初衷并非为了盈利，只是想帮助残疾群体。
Her friend, Zhang Quan, could only use WeChat to communicate with her even when they were sitting face to face, as Wang did not know sign language.
她的一位名叫张权的聋人朋友，由于王娜娜不懂手语，所以平时即使两人面对面坐在一起，张权也不得不用微信打字和她交流。
She felt sorry for Zhang when, surrounded by a chatting group, he could only remain silent.
当许多人一起聊天时，张权只能保持沉默，她就感到抱歉。
At the time, Wang was a computer science student studying image-based sign language recognition technology.
当时，计算机专业的王娜娜正在研究基于图像的手势识别技术。
However, she was too focused on technical details to consider the portability of an interpretation device. It was her close friend Huang Shuang who contributed the idea of creating something wearable.
不过，她本人将注意力过多放在考虑该翻译装置可携带性的相关技术细节之上，后来还是她的好朋友黄爽提醒她可以尝试发明可佩带的装置。
The idea for a sign language armband was inspired by a device used by student musicians studying piano.
手语臂环的想法，是受到学习钢琴的学生乐手们使用的一个装备的启发。
Teachers can tell whether the muscles in a student's arm are properly aligned through the use of special armbands.
老师可以通过这一特殊臂环的使用，分辨出学生胳膊上的肌肉是否均衡发力。
Wang and Huang decided to apply this technology to their device. They put each neuromuscular signal of sign language into a system, and then employed artificial intelligence to help the system recognize and "translate" the signs into audible information.
随后，娜娜和黄爽两人便决定将该项技术运用到自己的装置之上，她们将每个手语词汇的肌肉电信号录入系统，让人工智能深度学习，以便系统在下次遇见相应手语之后可进行识别和转化成语音。
In this way, people dependent on sign language to communicate can "speak" fluently with ordinary people.
通过这种方式，那些依赖手语交流的群体今后将有可能做到和正常人一样流畅“对话”。
