# Speech to text in Asterisk **URL:** \ **Category:** Asterisk Support\ **Created:** [May 9, 2017, 6:49am UTC](https://community.asterisk.org/t/speech-to-text-in-asterisk/70669 "2017-05-09T06:49:47Z")\ **Posts on this page:** 9\ **Page:** 1
**Author:** ![minkoko.gnext](https://sea1.discourse-cdn.com/asterisk/user_avatar/community.asterisk.org/minkoko.gnext/32/2082_2.png) [@minkoko.gnext](https://community.asterisk.org/u/minkoko.gnext)\ **Post date:** [May 9, 2017, 6:49am UTC](https://community.asterisk.org/t/speech-to-text-in-asterisk/70669/1 "2017-05-09T06:49:47Z")
Is asterisk support realtime speech to text transcription? Because I need speech recognition while phone conversation between call center and customer. for example ext 100:hello ext 200:hello I’m blah blah blah Is there anyway to transcript with Japanese language? Please give me advice. I need urgent. Thank you ---
**Author:** ![jcolp](https://sea1.discourse-cdn.com/asterisk/user_avatar/community.asterisk.org/jcolp/32/8848_2.png) [@jcolp](https://community.asterisk.org/u/jcolp)\ **Post date:** [May 9, 2017, 9:51am UTC](https://community.asterisk.org/t/speech-to-text-in-asterisk/70669/2 "2017-05-09T09:51:40Z")
Speech recognition and speech to text are two different things. People have written AGIs to use outside services for speech recognition, but I don’t think anyone has done speech to text. Regardless Asterisk itself does not provide an engine to do this. It’s up to something outside. ---
**Author:** ![johnkiniston](https://sea1.discourse-cdn.com/asterisk/user_avatar/community.asterisk.org/johnkiniston/32/28_2.png) [@johnkiniston](https://community.asterisk.org/u/johnkiniston)\ **Post date:** [May 9, 2017, 3:28pm UTC](https://community.asterisk.org/t/speech-to-text-in-asterisk/70669/3 "2017-05-09T15:28:33Z")
You could look at Clarify, [http://clarify.io/](http://clarify.io/) They had a product at Astricon a few years back that did word tagging from conversations via an API. ---
**Author:** ![minkoko.gnext](https://sea1.discourse-cdn.com/asterisk/user_avatar/community.asterisk.org/minkoko.gnext/32/2082_2.png) [@minkoko.gnext](https://community.asterisk.org/u/minkoko.gnext)\ **Post date:** [May 11, 2017, 2:45am UTC](https://community.asterisk.org/t/speech-to-text-in-asterisk/70669/4 "2017-05-11T02:45:00Z")
Thank you so much for your reply. ---
**Author:** ![jersonjunior](https://sea1.discourse-cdn.com/asterisk/user_avatar/community.asterisk.org/jersonjunior/32/1589_2.png) [@jersonjunior](https://community.asterisk.org/u/jersonjunior)\ **Post date:** [May 11, 2017, 8:05pm UTC](https://community.asterisk.org/t/speech-to-text-in-asterisk/70669/5 "2017-05-11T20:05:11Z")
Use IBM Speech to Text is easy to implement with asterisk! [https://www.ibm.com/watson/developercloud/speech-to-text.html](https://www.ibm.com/watson/developercloud/speech-to-text.html) ---
**Author:** ![minkoko.gnext](https://sea1.discourse-cdn.com/asterisk/user_avatar/community.asterisk.org/minkoko.gnext/32/2082_2.png) [@minkoko.gnext](https://community.asterisk.org/u/minkoko.gnext)\ **Post date:** [May 12, 2017, 1:51am UTC](https://community.asterisk.org/t/speech-to-text-in-asterisk/70669/6 "2017-05-12T01:51:35Z")
I used IBM Speech to Text,but it is not OK for Speaker label.Please suggest me. Thank jersonjunior for your suggestion. ---
**Author:** ![ianplain](https://sea1.discourse-cdn.com/asterisk/user_avatar/community.asterisk.org/ianplain/32/1776_2.png) [@ianplain](https://community.asterisk.org/u/ianplain)\ **Post date:** [May 12, 2017, 3:18pm UTC](https://community.asterisk.org/t/speech-to-text-in-asterisk/70669/7 "2017-05-12T15:18:45Z")
We have done voicemail to text with googles API , quality is varible due to teh way people speak and how googles spech to text system works. [http://www.cyber-cottage.eu/?p=1801](http://www.cyber-cottage.eu/?p=1801) gives brief overview email for more info ---
**Author:** ![poing](https://sea1.discourse-cdn.com/asterisk/user_avatar/community.asterisk.org/poing/32/1696_2.png) [@poing](https://community.asterisk.org/u/poing)\ **Post date:** [May 13, 2017, 2:58am UTC](https://community.asterisk.org/t/speech-to-text-in-asterisk/70669/8 "2017-05-13T02:58:20Z")
I’ve done some basic work with Google Speech API, even Japanese. You can find a working prototype here [GitHub](https://github.com/invitecomm/asterisk-ivr/blob/master/pigeonhole/speech-recog.agi). You will run into some serious complications trying to translate a long conversation. I use it for a IVR system, asking and translating response questions for a survey. 弊社オフィスは市ヶ谷にございます。デモをお見せすることもできますので、いつでもお立ち寄りください。 ---
**Author:** ![david551](https://avatars.discourse-cdn.com/v4/letter/d/3bc359/32.png) [@david551](https://community.asterisk.org/u/david551)\ **Post date:** [May 13, 2017, 10:40am UTC](https://community.asterisk.org/t/speech-to-text-in-asterisk/70669/9 "2017-05-13T10:40:32Z")
Speech to text is starting off at a big disadvantage on telephones anyway, because the high frequency response of the PSTN, and codecs designed for it, doesn’t accurately reproduce sibilants (s, h, sh, etc.). Also, in my view,speaker independent, continuous, speech recognition, by machine, is still a research topic.