1 00:00:00,300 --> 00:00:02,940 ‫So now let's talk about Amazon Transcribe. 2 00:00:02,940 --> 00:00:04,170 ‫So as the name indicates, 3 00:00:04,170 --> 00:00:07,980 ‫it allows you to automatically convert speech into text. 4 00:00:07,980 --> 00:00:10,440 ‫So you pass in some audio 5 00:00:10,440 --> 00:00:13,140 ‫and automatically is going to be transcribed into text, 6 00:00:13,140 --> 00:00:14,850 ‫so you could say, hey, hello, 7 00:00:14,850 --> 00:00:18,120 ‫my name is Stephane and I hope you're enjoying the course. 8 00:00:18,120 --> 00:00:18,953 ‫Now, how does it do it? 9 00:00:18,953 --> 00:00:22,050 ‫Well, it uses a deep learning process called the ASR, 10 00:00:22,050 --> 00:00:24,600 ‫the Automatic Speech Recognition, 11 00:00:24,600 --> 00:00:28,560 ‫to convert speech to text very quickly and accurately 12 00:00:28,560 --> 00:00:30,150 ‫and there are a few features you need to know about. 13 00:00:30,150 --> 00:00:34,350 ‫The first one is that you can automatically remove any PII, 14 00:00:34,350 --> 00:00:38,400 ‫so personally identifiable information, using reduction. 15 00:00:38,400 --> 00:00:39,450 ‫So that means, for example, 16 00:00:39,450 --> 00:00:43,440 ‫if you have someone's age or name or Social Security Number, 17 00:00:43,440 --> 00:00:45,660 ‫this can be automatically removed 18 00:00:45,660 --> 00:00:47,820 ‫and also you can have access 19 00:00:47,820 --> 00:00:52,020 ‫to automatic language identification for multilingual audio. 20 00:00:52,020 --> 00:00:53,850 ‫So if you have some French and some English, 21 00:00:53,850 --> 00:00:56,010 ‫some Spanish, Transcribe is smart enough 22 00:00:56,010 --> 00:00:58,530 ‫to recognize all of those. 23 00:00:58,530 --> 00:01:00,690 ‫So the use cases for Amazon Transcribe 24 00:01:00,690 --> 00:01:03,000 ‫is to transcribe customer service calls, 25 00:01:03,000 --> 00:01:05,760 ‫to automate closed captioning and subtitling, 26 00:01:05,760 --> 00:01:07,620 ‫and also to generate metadata 27 00:01:07,620 --> 00:01:11,223 ‫for media assets to create a fully searchable archive. 28 00:01:12,210 --> 00:01:14,850 ‫So let's have a look at Amazon Transcribe. 29 00:01:14,850 --> 00:01:18,120 ‫So I can create a transcript and I can see 30 00:01:18,120 --> 00:01:21,210 ‫that the specific language right now is set to English US, 31 00:01:21,210 --> 00:01:23,060 ‫so now if I click on start streaming, 32 00:01:24,060 --> 00:01:26,073 ‫hello, I really like this course. 33 00:01:27,960 --> 00:01:30,360 ‫As you can see, the outcome of the audio 34 00:01:30,360 --> 00:01:32,910 ‫get directly transcribed into some text, 35 00:01:32,910 --> 00:01:34,350 ‫so that's pretty cool, right? 36 00:01:34,350 --> 00:01:37,380 ‫Also, you can have the setting to remove some content 37 00:01:37,380 --> 00:01:39,780 ‫and you can remove PII identification. 38 00:01:39,780 --> 00:01:41,310 ‫So I could say, hey, 39 00:01:41,310 --> 00:01:44,340 ‫you can identify and redact it to remove it 40 00:01:44,340 --> 00:01:46,770 ‫and these are the kind of things that can be removed. 41 00:01:46,770 --> 00:01:48,930 ‫So for example, let's have an example 42 00:01:48,930 --> 00:01:53,850 ‫and try it out, so I won't show this all again. 43 00:01:53,850 --> 00:01:57,870 ‫Hello, my name is Stephane, I am 31 years old 44 00:01:57,870 --> 00:02:02,870 ‫and my phone number is 910-747-280 45 00:02:07,607 --> 00:02:09,930 ‫and as you can see now, the things got hidden. 46 00:02:09,930 --> 00:02:12,930 ‫So my name is hidden and my phone is hidden. 47 00:02:12,930 --> 00:02:15,300 ‫Obviously this was not my real phone number. 48 00:02:15,300 --> 00:02:16,590 ‫You cannot try it. 49 00:02:16,590 --> 00:02:18,510 ‫Okay and last thing I want to show you 50 00:02:18,510 --> 00:02:21,180 ‫is that you can actually stream into multiple languages. 51 00:02:21,180 --> 00:02:23,760 ‫So you have automatic language identification. 52 00:02:23,760 --> 00:02:26,400 ‫So I would choose English and French 53 00:02:26,400 --> 00:02:27,843 ‫and then let's start again. 54 00:02:31,140 --> 00:02:34,503 ‫Hello, this is some recognition happening in English. 55 00:02:37,508 --> 00:02:40,591 ‫(speaking in French) 56 00:02:45,600 --> 00:02:46,920 ‫Pretty awesome, right? 57 00:02:46,920 --> 00:02:48,840 ‫Well, that's it for Amazon Transcribe. 58 00:02:48,840 --> 00:02:50,250 ‫If you wanted to play some more, 59 00:02:50,250 --> 00:02:52,470 ‫you can have a look at all the options in the bottom, 60 00:02:52,470 --> 00:02:53,640 ‫but that's it for this lecture. 61 00:02:53,640 --> 00:02:56,733 ‫I hope you liked it and I will see you in the next lecture.