<html>
  <head>
    <meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
  </head>
  <body>
    <font face="monospace">Hi Mark,<br>
      <br>
      But using the media_exchange you can "fork" (as a new SIP call)
      only one of the RTP streams - of course, the TTS engine should be
      able to accept pure SIP calls.<br>
      <br>
      Regards,<br>
    </font>
    <pre class="moz-signature" cols="72">Bogdan-Andrei Iancu

OpenSIPS Founder and Developer
  <a class="moz-txt-link-freetext" href="https://www.opensips-solutions.com">https://www.opensips-solutions.com</a>
OpenSIPS eBootcamp 2021 
  <a class="moz-txt-link-freetext" href="https://opensips.org/training/OpenSIPS_eBootcamp_2021/">https://opensips.org/training/OpenSIPS_eBootcamp_2021/</a></pre>
    <div class="moz-cite-prefix">On 9/17/21 3:50 PM, Mark Allen wrote:<br>
    </div>
    <blockquote type="cite"
cite="mid:CADaqb1sp+oOCFQc0kQSa6KTNZ-PKXvObmd_8pir7ij5u3s8hUw@mail.gmail.com">
      <meta http-equiv="content-type" content="text/html; charset=UTF-8">
      <div dir="ltr">Thanks for that Johan - I hadn't thought about that
        aspect. All theoretic at the moment, but IBM Voice Gateway, at
        least, does claim to be able to handle it using SIPREC - so
        maybe they are confident about their ability to differentiate
        between caller and callee in a single stream?...
        <div><br>
        </div>
        <blockquote class="gmail_quote" style="margin:0px 0px 0px
          0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">"The
          voice gateway provides the ability to transcribe caller and
          callee (e.g. contact-center agent) audio from an active phone
          call in real time using the SIPREC protocol." - <a
href="https://www.ibm.com/docs/en/voice-gateway?topic=gateway-about-voice"
            moz-do-not-send="true">https://www.ibm.com/docs/en/voice-gateway?topic=gateway-about-voice</a><br>
        </blockquote>
        <div> </div>
      </div>
      <br>
      <div class="gmail_quote">
        <div dir="ltr" class="gmail_attr">On Fri, 17 Sept 2021 at 10:33,
          johan <<a href="mailto:johan@democon.be"
            moz-do-not-send="true">johan@democon.be</a>> wrote:<br>
        </div>
        <blockquote class="gmail_quote" style="margin:0px 0px 0px
          0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">
          <div>
            <p>The issue with siprec (based on rtpproxy) is that you
              have only 1 stream containing the voice from caller to
              callee and callee to caller. So that will give a hard time
              on the ASR :-).  I do know that rtpengine has something
              similar to siprec but I don't know the details. <br>
            </p>
            <p><br>
            </p>
            <p>Bottom line, in my opinion, you need to have 2 separate
              streams before you can start STT. <br>
            </p>
            <p><br>
            </p>
            <p>wkr, <br>
            </p>
            <p><br>
            </p>
            <div>On 17/09/2021 11:04, Mark Allen wrote:<br>
            </div>
            <blockquote type="cite">
              <div dir="ltr">I'm just starting to look at Speech-to-Text
                (STT) processing for calls - initially recordings but
                moving on to real-time. I would see this working along
                the lines of either: 
                <div><br>
                </div>
                <div>- a call is recorded, and when the call ends an
                  event is triggered to initiate transcription of the
                  recording</div>
                <div>- a call starts, the RTP is forked to the STT
                  engine which sends real-time transcription<br>
                  <div><br>
                  </div>
                  <div>I can see that with OpenSIPS, the SIPREC and
                    Media Exchange modules allow for forking of the RTP,
                    providing a means of sending the data for
                    processing, but is anybody actually doing this? If
                    so, what has been your experience? Is there a
                    toolset that works well with this (e.g. IBM Voice
                    Gateway, Google, Amazon etc)? </div>
                </div>
              </div>
              <br>
              <fieldset></fieldset>
              <pre>_______________________________________________
Users mailing list
<a href="mailto:Users@lists.opensips.org" target="_blank" moz-do-not-send="true">Users@lists.opensips.org</a>
<a href="http://lists.opensips.org/cgi-bin/mailman/listinfo/users" target="_blank" moz-do-not-send="true">http://lists.opensips.org/cgi-bin/mailman/listinfo/users</a>
</pre>
            </blockquote>
          </div>
          _______________________________________________<br>
          Users mailing list<br>
          <a href="mailto:Users@lists.opensips.org" target="_blank"
            moz-do-not-send="true">Users@lists.opensips.org</a><br>
          <a
            href="http://lists.opensips.org/cgi-bin/mailman/listinfo/users"
            rel="noreferrer" target="_blank" moz-do-not-send="true">http://lists.opensips.org/cgi-bin/mailman/listinfo/users</a><br>
        </blockquote>
      </div>
      <br>
      <fieldset class="mimeAttachmentHeader"></fieldset>
      <pre class="moz-quote-pre" wrap="">_______________________________________________
Users mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Users@lists.opensips.org">Users@lists.opensips.org</a>
<a class="moz-txt-link-freetext" href="http://lists.opensips.org/cgi-bin/mailman/listinfo/users">http://lists.opensips.org/cgi-bin/mailman/listinfo/users</a>
</pre>
    </blockquote>
    <br>
  </body>
</html>