Error while doing Continuous recognition (speech to text)

[CN-Paul]

[901014]: 70090ms SPX_DBG_TRACE_VERBOSE:  usp_reco_engine_adapter.cpp:1163 speech.config {"context":{"system":{"version":"1.37.0","name":"SpeechSDK","build":"Linux-x64"},"os":{"name":"Linux 6.1.84-99.169.amzn2023.x86_64 #1 SMP PREEMPT_DYNAMIC Mon Apr  8 19:19:48 UTC 2024 x86_64","version":"6.1.84-99.169.amzn2023.x86_64","platform":"Linux"},"audio":{"source":{"type":"Stream","model":"","samplerate":"16000","bitspersample":"16","channelcount":"1"}}}}
[901014]: 70090ms SPX_TRACE_INFO:  usp_connection.cpp:787 Create requestId  for messageType 0
[901014]: 70090ms SPX_DBG_TRACE_SCOPE_EXIT:  usp_reco_engine_adapter.cpp:230 SetFormat
[901014]: 70090ms SPX_DBG_TRACE_SCOPE_EXIT:  audio_stream_session.cpp:3397 InformAdapterSetFormatStarting
[901014]: 70090ms SPX_TRACE_INFO:  web_socket.cpp:800 DoWork: open transport.
[901014]: 70090ms SPX_TRACE_INFO:  web_socket.cpp:479 Start to open websocket. WebSocket: 0x9c020310, wsio handle: 0x9c01ff50
[901014]: 70090ms SPX_TRACE_SCOPE_ENTER:  uws_web_socket.cpp:149 Open
[901014]: 70090ms SPX_TRACE_INFO: AZ_LOG_INFO:  tlsio_openssl.c:2027 create_openssl_instance by TLS_method.
[901014]: 70091ms SPX_TRACE_INFO: AZ_LOG_INFO:  tlsio_openssl.c:1849 load_system_store not implemented on this platform
[901014]: 70091ms SPX_TRACE_INFO: AZ_LOG_INFO:  tlsio_openssl.c:1878 CRL check off, as requested.
[901014]: 111557ms SPX_TRACE_ERROR: AZ_LOG_ERROR:  tlsio_openssl.c:691 error:14200010:SSL routines:func(512):EC lib
[901014]: 111557ms SPX_TRACE_SCOPE_EXIT:  uws_web_socket.cpp:149 Open
[901014]: 112167ms SPX_TRACE_ERROR: AZ_LOG_ERROR:  tlsio_openssl.c:2464 FORCE-Closing tlsio instance.

trying to make a zoom bot that does speech to text using speech sdk(c++) and zoom meetings sdk. I've changed my openssl version to 1.1.1l, but still sufferring openssl error.

I've also tried https://learn.microsoft.com/en-us/azure/ai-services/speech-service/how-to-configure-openssl-linux?pivots=programming-language-cpp but still having error log above.

When I debug the program, I'm having segmentation fault at code after

recognizer->StartContinuousRecognitionAsync().wait();

recognitionEnd.get_future().wait();

is called.

Can anyone suggest any potential problems that I migth be missing? I'm suffering with this problem for weeks.

[github-actions[bot]]

This item has been open without activity for 19 days. Provide a comment on status and remove "update needed" label.

[tamagoeggo]

@CN-Paul, for the meeting bot portion, you could check out the Recall.ai API for your meeting bots instead. It’s an API that lets you use meeting bots to get audio, video, transcripts and metadata from meetings without you needing to spend months to build, scale and maintain these bots.

Disclaimer: I’m one of the founders.

[github-actions[bot]]

This item has been open without activity for 19 days. Provide a comment on status and remove "update needed" label.

[pankopon]

"version":"6.1.84-99.169.amzn2023.x86_64"

Amazon Linux 2023 is not officially a supported Linux distribution at the moment, but it has been tested to work with the Speech SDK in C++ when using the Amazon Linux 2023 Docker image and the following setup:

yum updateinfo && yum update -y
yum install -y alsa-lib ca-certificates openssl wget
yum groupinstall -y "Development tools"

Use the latest Speech SDK release (1.40.0 as of this writing) and do not install any additional OpenSSL versions or set SSL environment variables. The default OpenSSL version in Amazon Linux 2023 will work as is.

Also note that since the 1.38.0 release, all Speech SDK packages include the following 2 new library files that are required for the OpenSSL version detection to work:

libpal_azure_c_shared.so
libpal_azure_c_shared_openssl3.so

If your project is set up so that SDK libraries from the release packages are copied to a custom location, make sure that these libpal libraries are copied as well, otherwise you will have errors. All Speech SDK libraries must reside in the same directory.

Please re-test based on the above. If you still have issues then open a new bug report with full details.