月球采矿
10 Second Trivia. What would you most likely find on the surface of the moon? Regolith, Dark Side, Liquid Water or Sea of Placidity. Regolith, also known as lunar soil or moon dust is all over the moon.
Of course, regolith is all over the earth too but that`s not the kind the European Space Agency wants to start mining. It recently said it was teaming up with the European Aerospace Company to look into mining lunar soil. Why? Lunar regolith, like regolith on earth has water and oxygen in it but being able to mine those elements on the moon could help support a potential base there. And if scientists are also able to extract helium 3, an isotope that`s believe to be in lunar soil they think that could be used to develop rocket fuel on the moon.
All that would add up to using the moon as a jumping off point to exploring deeper into space. Scientists hope to get this mission off the ground by the year 2025 but there are some obstacles to overcome. For one thing, the rocket that would carry mining equipment to the moon hasn`t been built yet.
It`s currently under development. Also, the lander that would actually place the equipment on the moon still has to be made. So the European Space Agency is really just taking the first step here to see if mining the moon is possible and if it makes sense in terms of funding and the resources it would require. Still, according to Fortune.com China and India are also looking into the moon as a possible source of helium 3. So there could be an international competition to mine the moon taking place about 240,000 miles away from it.
智能识别
For decades, the film industries had the technology to alter picture or sound to make something that never happened look like it really did. A great example is when the movie character Forrest Gump appeared to meet the real life President Lyndon Johnson but this wasn`t easy to do. It took a lot of time, a lot of skill, a lot of money. Because of advancements in artificial intelligence though, it`s getting easier and cheaper to manipulate audio or video and the government and the military are concerned this could be used to spread false information. Here`s a look at what they`re doing to separate fact from fiction. (inaudible) manipulate people`s voices. It`s actually quite difficult and people are very good at picking this up. When it does fool people is when it`s in a lot of noise. You create a noisy environment where you can`t hear things very clearly. Then it becomes difficult to tell that this is fake and that`s where you really need these automatic techniques which can learn to ignore the noise. You`re building this technology. How would you guys be able to detect this?
UNIDENTIFIED MALE: For anybody that`s enrolled, by enrolled we mean we`ve taught the machine what the person looks like and what they sound like. We use voice recognition combined with this face recognition. So in this case, it shows when the face recognizer thinks it knows who this is, Chris in this case and it also shows when it thinks the voice. This is one of your colleagues, Chris? Yes. What if somebody sounds really like Chris? So I`ve got a case where it`s not far off from Chris and they try to align the words so that that`s similar. So it recognizes Chris` face but it recognizes somebody else`s voice. So it says, hey there`s an inconsistency here. So here, I show a little bit more about how we track the mouth to check to make sure that the motions of the mouth correspond to the words being spoken. The line down at the bottom says there`s a lot of problem with the lip sync over most of the video.
So what if there`s a video. It`s Chris` real voice but it`s just been cut up to make it look like he`s said something here that he said previously. Yes. So there`d be a change in lip sync. There`d also be some pretty obvious artifacts that even if they aren`t obvious to your ear would be obvious in the audio. That it`s not just a recording. It`s a synthesis. One of the things that differentiates what we are doing from - - from a lot of others is rather than just focusing on a single modality be it imagery or video or audio. We`re looking at how those interact.
自动停车
CARL AZUZ: At Gatwick Airport in London, UK a new type of parking attendant is making headlines for being completely robotic. It`s name is Stan. It looks a little like a miniature flatbed truck but when a car is ready to be parked or picked up, Stan shows up, hoists it up like a forklift and then ferries it to its space. The company that makes the robot says it can create up to 50 percent more spaces for cars. Apparently by parking them closer together.
月球开矿停车 注释
1. regolith 风化层：基岩上面的松散的岩石表层，组成大多数土地的表面/'regəlɪθ/
2. Sea of Placidity静海/plə'sidəti/
3. European Space Agency欧洲航天局
4. team up with与…合作
5. aerospace航空航天技术/ˈeə.rəʊ.speɪs/
6. look into研究
7. mine开采
8. element元素
9. base基地
10. extract提取
11. helium 3氦三同位素 /'hiːlɪəm/
12. isotope同位素/'aɪsətəʊp/
13. rocket fuel火箭燃料
14. add up to意味着
15. get......off the ground 启动, 使取得进展
16. obstacle障碍
17. For one thing首先
18. under development处于研发阶段
19. lander 登陆器
20. make sense讲得通；言之有理
21. Fortune.com据福布斯网站
22. alter改变，更改
23. Forrest Gump《阿甘正传》的阿甘
24. real life现实生活（的）
25. Lyndon Johnson林登·约翰逊（第三十六任美国总统）
26. advancement进展
27. manipulate操纵
28. audio音频
29. concerned担忧的
30. fiction虚构，捏造
31. pick up识别
32. fool欺骗
33. automatic自动的
34. ignore忽视
35. detect发现
36. enroll登记
37. recognition识别
38. recognizer识别器
39. far off from远离
40. align the words对口型，音像匹配
41. inconsistency不一致
42. track追踪，检测
43. motion动作
44. correspond to相当于…，符合
45. down at the bottom在底端
46. lip sync对口型（lip synchronization）
47. cut up剪切
48. artifact/'ɑ:təˌfækt/p图痕迹
49. recording录像
50. synthesis /'sɪnθɪsɪs/合成品，此指p视频
51. differentiate区分 /,dɪfə'renʃɪeɪt/
52. modality形态 /mə(ʊ)'dælɪtɪ/
53. be it A or B 无论是A 还是B
54. imagery画像；照片 /ˈɪmɪdʒəri/
55. Gatwick Airport盖特威克机场,伦敦第二大的机场
56. parking attendant泊车员; 看车人
57. make headlines成为头条新闻
58. robotic自动的
59. Stan斯坦，法国初创Stanley Robotics设计开发的一款泊车机器人
60. miniature微型的/'mɪnɪtʃə/
61. flatbed truck平板货车
62. hoist up升起
63. forklift叉车
64. ferry渡运; 运送
65. apparently显然
联系客服
