1 00:00:25,890 --> 00:00:16,960 monitor to alpha Bravo 2 00:00:25,900 --> 00:00:29,000 change rate 3 00:00:34,010 --> 00:00:31,550 didn't recognize it it won't play okay 4 00:00:40,360 --> 00:00:34,020 so the Winx I know that it took the 5 00:00:44,049 --> 00:00:40,370 change down 6 00:00:44,059 --> 00:00:48,930 change rate 7 00:00:54,060 --> 00:00:51,979 stop Marcus are you probably marking us 8 00:01:00,110 --> 00:00:54,070 but I'm gonna mark some of the ones that 9 00:01:00,120 --> 00:01:03,099 go 10 00:01:03,109 --> 00:01:13,109 zoom in 11 00:01:13,119 --> 00:01:15,910 right 12 00:01:20,770 --> 00:01:18,830 change rate 13 00:01:25,550 --> 00:01:20,780 right 14 00:01:41,390 --> 00:01:27,560 I think I would cycle a whole thing one 15 00:01:50,110 --> 00:01:43,770 okay 16 00:01:55,800 --> 00:01:50,120 monitor one monitor one monitor two 17 00:02:00,210 --> 00:01:58,219 standby 18 00:02:06,370 --> 00:02:00,220 activate 19 00:02:10,279 --> 00:02:06,380 go 20 00:02:10,289 --> 00:02:19,839 right stop 21 00:02:30,460 --> 00:02:22,170 change rate 22 00:02:36,400 --> 00:02:32,320 easy 23 00:02:38,690 --> 00:02:36,410 so cameras 24 00:02:40,640 --> 00:02:38,700 no cameras 25 00:02:44,290 --> 00:02:40,650 more 26 00:02:51,680 --> 00:02:44,300 too much 27 00:02:51,690 --> 00:02:59,559 change rate 28 00:02:59,569 --> 00:03:10,460 tilt up 29 00:03:10,470 --> 00:03:15,160 right 30 00:03:22,890 --> 00:03:17,120 stop 31 00:03:27,010 --> 00:03:25,030 tilt up 32 00:03:39,280 --> 00:03:27,020 stop 33 00:03:41,380 --> 00:03:39,290 select one time ideally the mood active 34 00:03:54,099 --> 00:03:41,390 LED is on the stamp a lit is also I beat 35 00:03:55,720 --> 00:03:54,109 you back the baby so if we I'll talk to 36 00:03:56,849 --> 00:03:55,730 chef about how we're gonna manage that 37 00:03:59,349 --> 00:03:56,859 okay 38 00:04:02,080 --> 00:03:59,359 like I told chef what's most important 39 00:04:04,089 --> 00:04:02,090 to that are most affects your opinion 40 00:04:09,250 --> 00:04:04,099 sure that is our most important feedback 41 00:04:14,360 --> 00:04:11,360 this would be where you'd be activated 42 00:04:23,050 --> 00:04:14,370 if you wanted to or as required that's 43 00:04:27,770 --> 00:04:25,970 but the data is compared against ground 44 00:04:30,590 --> 00:04:27,780 data that we acquired during astronaut 45 00:04:32,720 --> 00:04:30,600 training hopefully we'll be able to see 46 00:04:35,150 --> 00:04:32,730 are the effects the space environment 47 00:04:38,620 --> 00:04:35,160 has on the speech process and voice 48 00:04:44,930 --> 00:04:40,970 there are several successful 49 00:04:47,290 --> 00:04:44,940 applications of voice commanding systems 50 00:04:50,150 --> 00:04:47,300 in in the commercial sector for example 51 00:04:52,520 --> 00:04:50,160 in the area of aid to the handicapped 52 00:04:56,390 --> 00:04:52,530 and handicapped person can control by 53 00:04:59,780 --> 00:04:56,400 voice a motorized wheelchair medical 54 00:05:03,170 --> 00:04:59,790 reporting currently doctors are using 55 00:05:04,880 --> 00:05:03,180 voice recognition to to generate medical 56 00:05:07,880 --> 00:05:04,890 reports whereby they're looking at a 57 00:05:09,710 --> 00:05:07,890 x-ray for example and are examining it 58 00:05:11,660 --> 00:05:09,720 and are in putting words into the 59 00:05:13,760 --> 00:05:11,670 recognizer that it recognizes in its 60 00:05:16,700 --> 00:05:13,770 generating report about the doctor's 61 00:05:18,820 --> 00:05:16,710 examination other applications is access 62 00:05:21,470 --> 00:05:18,830 control where they use a speaker 63 00:05:23,210 --> 00:05:21,480 verification system is the person that 64 00:05:25,730 --> 00:05:23,220 wants to get into this door truly the 65 00:05:27,860 --> 00:05:25,740 person it is and they use a specific 66 00:05:29,930 --> 00:05:27,870 voice print algorithm to try and 67 00:05:33,620 --> 00:05:29,940 determine if this person is indeed the 68 00:05:35,540 --> 00:05:33,630 person he says he is the system that we 69 00:05:38,150 --> 00:05:35,550 use is what they call speaker dependence 70 00:05:40,070 --> 00:05:38,160 system which means that the crew each 71 00:05:42,410 --> 00:05:40,080 crew person has to train the system 72 00:05:44,570 --> 00:05:42,420 before it can use it training consists 73 00:05:46,910 --> 00:05:44,580 of going through our vocabulary times 74 00:05:49,310 --> 00:05:46,920 it's saying its word each word is then 75 00:05:51,230 --> 00:05:49,320 digitized and stored in memory as a 76 00:05:53,750 --> 00:05:51,240 template or voice print similar to a 77 00:05:55,880 --> 00:05:53,760 person's thumbprint and then all those 78 00:05:59,010 --> 00:05:55,890 words are then saved according to each 79 00:06:03,129 --> 00:06:02,589 okay when the astronaut wants to use the 80 00:06:04,330 --> 00:06:03,139 system 81 00:06:06,999 --> 00:06:04,340 he would come in you know plug in his 82 00:06:08,740 --> 00:06:07,009 headset and plug in the display in the 83 00:06:11,589 --> 00:06:08,750 shorting plug and then all he has to do 84 00:06:14,249 --> 00:06:11,599 is turn on the power switch to on and 85 00:06:17,649 --> 00:06:14,259 that's the only manual step necessary 86 00:06:19,119 --> 00:06:17,659 after when the system powers up then the 87 00:06:21,279 --> 00:06:19,129 system automatically krons 88 00:06:23,649 --> 00:06:21,289 yes not to say his name when it says his 89 00:06:25,420 --> 00:06:23,659 name it recognizes who he is would go 90 00:06:28,180 --> 00:06:25,430 out and load his templates from memory 91 00:06:30,219 --> 00:06:28,190 into the recognizer then when he wants 92 00:06:32,529 --> 00:06:30,229 to control a system you'll say a command 93 00:06:34,180 --> 00:06:32,539 it takes that word and digitizes it and 94 00:06:36,279 --> 00:06:34,190 compares it with the words that were 95 00:06:38,559 --> 00:06:36,289 loaded into the recognizer if it finds a 96 00:06:40,719 --> 00:06:38,569 match it sends that infrared sensing 97 00:06:42,520 --> 00:06:40,729 information from the recognizer to are 98 00:06:44,559 --> 00:06:42,530 controlling Hardware where we process 99 00:06:47,110 --> 00:06:44,569 information figure out what word it was 100 00:06:49,089 --> 00:06:47,120 and then send the appropriate signal out 101 00:06:50,749 --> 00:06:49,099 to the CCTV system to control the camera 102 00:06:53,719 --> 00:06:50,759 or movement or 103 00:06:55,790 --> 00:06:53,729 whatever the function happens the voice 104 00:06:57,920 --> 00:06:55,800 command system is the first experiment 105 00:07:00,230 --> 00:06:57,930 to be flown in space which uses voices 106 00:07:02,299 --> 00:07:00,240 to control we think this is going to be 107 00:07:04,640 --> 00:07:02,309 useful to be astronauts for two reasons 108 00:07:06,409 --> 00:07:04,650 and the first reason is that by having a 109 00:07:08,179 --> 00:07:06,419 system which is voice controlled the 110 00:07:11,179 --> 00:07:08,189 astronauts hands and eyes are free to 111 00:07:13,519 --> 00:07:11,189 perform other tasks the second reason is 112 00:07:15,200 --> 00:07:13,529 that by using a voice control system it 113 00:07:17,510 --> 00:07:15,210 isn't necessary for you to be in the 114 00:07:18,980 --> 00:07:17,520 location where the system is in order to 115 00:07:23,929 --> 00:07:18,990 control it so you're free to move around 116 00:07:26,029 --> 00:07:23,939 and you can be remotely located no we 117 00:07:28,399 --> 00:07:26,039 can have as many people as want to can 118 00:07:32,329 --> 00:07:28,409 use our system we have it set up so that 119 00:07:34,610 --> 00:07:32,339 only five five people can actually have 120 00:07:37,279 --> 00:07:34,620 their voices put on to e-square bumps 121 00:07:40,100 --> 00:07:37,289 but we also have the ability to retrain 122 00:07:41,689 --> 00:07:40,110 templates real-time so anybody who wants 123 00:07:43,760 --> 00:07:41,699 to use the system may not already have 124 00:07:45,320 --> 00:07:43,770 theirs templates made but they can come 125 00:07:47,389 --> 00:07:45,330 up to the system and make their template 126 00:07:48,679 --> 00:07:47,399 anytime they want so we're not limited 127 00:07:58,370 --> 00:07:48,689 on the number of people who can use the 128 00:08:02,170 --> 00:08:00,500 all right well we have 40 commands that 129 00:08:04,580 --> 00:08:02,180 are possible on the voice command system 130 00:08:07,070 --> 00:08:04,590 most of these commands are exact 131 00:08:09,230 --> 00:08:07,080 duplicates to what are on a CCTV manual 132 00:08:11,240 --> 00:08:09,240 switch panel these are selecting 133 00:08:12,830 --> 00:08:11,250 monitors cameras and working in the 134 00:08:16,010 --> 00:08:12,840 camera motions like panning and tilting 135 00:08:18,350 --> 00:08:16,020 zooming focusing all of those we also 136 00:08:20,810 --> 00:08:18,360 have some VCS unique commands which are 137 00:08:23,600 --> 00:08:20,820 used to activate the system deactivate 138 00:08:26,690 --> 00:08:23,610 the system identify the user and we also 139 00:08:29,510 --> 00:08:26,700 have some special macro commands for 140 00:08:31,160 --> 00:08:29,520 example we have a command that will stow 141 00:08:33,200 --> 00:08:31,170 all of the cameras in the payload Bay 142 00:08:35,270 --> 00:08:33,210 into the position that they have to be 143 00:08:38,300 --> 00:08:35,280 in for launch and entry simply by saying 144 00:08:40,520 --> 00:08:38,310 still cameras so we also have some 145 00:08:45,750 --> 00:08:40,530 special features that the manual switch 146 00:08:48,930 --> 00:08:47,699 well the way we chose the words to use 147 00:08:51,360 --> 00:08:48,940 is because they're the words that are 148 00:08:52,650 --> 00:08:51,370 actually on the switch panel so they're 149 00:08:55,410 --> 00:08:52,660 the ones that are familiar with when 150 00:08:58,199 --> 00:08:55,420 they think of moving a camera they think 151 00:09:00,629 --> 00:08:58,209 of tilting it up so we selected all of 152 00:09:02,900 --> 00:09:00,639 those works now they did tailor their 153 00:09:04,620 --> 00:09:02,910 vocabulary to their personal and use 154 00:09:06,509 --> 00:09:04,630 words that they were the most 155 00:09:07,319 --> 00:09:06,519 comfortable with words that work the 156 00:09:09,870 --> 00:09:07,329 best for them 157 00:09:11,370 --> 00:09:09,880 so both of our astronauts have have a 158 00:09:13,379 --> 00:09:11,380 different vocabulary but they're also 159 00:09:15,090 --> 00:09:13,389 very similar and they're the words that 160 00:09:24,370 --> 00:09:15,100 make sense to them and that's how we 161 00:09:29,260 --> 00:09:26,590 well there's not a lot known about the 162 00:09:31,660 --> 00:09:29,270 effects of microgravity on the voice we 163 00:09:33,790 --> 00:09:31,670 do know that in zero gravity that the 164 00:09:36,820 --> 00:09:33,800 position and the shape of the lungs and 165 00:09:39,550 --> 00:09:36,830 diaphragm are altered now this cause is 166 00:09:41,770 --> 00:09:39,560 a great enough distortion in the voice 167 00:09:44,590 --> 00:09:41,780 good recognition could be difficult to 168 00:09:49,720 --> 00:09:44,600 get so we're expecting that this may be 169 00:09:51,550 --> 00:09:49,730 a problem well what we're going to do is 170 00:09:53,860 --> 00:09:51,560 that if the astronauts do find that they 171 00:09:55,660 --> 00:09:53,870 have trouble getting recognition we have 172 00:09:58,300 --> 00:09:55,670 the ability to remake their templates 173 00:10:00,190 --> 00:09:58,310 real time so simply by clicking this 174 00:10:01,780 --> 00:10:00,200 mode switch try the Fitzsimmons 175 00:10:04,270 --> 00:10:01,790 automatically put into the template 176 00:10:06,220 --> 00:10:04,280 making mode and our system will prompt 177 00:10:08,020 --> 00:10:06,230 them through the entire vocabulary list 178 00:10:09,490 --> 00:10:08,030 they'll say each word twice and they'll 179 00:10:11,050 --> 00:10:09,500 have a brand new set of templates that 180 00:10:15,850 --> 00:10:11,060 are real and representative of their 181 00:10:17,050 --> 00:10:15,860 voice at that time it feels so fabulous 182 00:10:18,850 --> 00:10:17,060 you know there's times when you really 183 00:10:21,220 --> 00:10:18,860 think it isn't gonna fly it's just it's 184 00:10:22,930 --> 00:10:21,230 not gonna go and now not only is it 185 00:10:25,120 --> 00:10:22,940 going to go but our astronauts love it 186 00:10:26,920 --> 00:10:25,130 they love it they're having fun with it 187 00:10:29,320 --> 00:10:26,930 everybody we've shown it to like that 188 00:10:31,240 --> 00:10:29,330 people are coming up out of the corners 189 00:10:32,290 --> 00:10:31,250 with ideas of how to use it and it