I am using Azure Speech SDK
for speech recognition in my application (chrome extension).
The problem I am facing now is, if I am not using a headset, then the speech from another person (echo) in the conference is getting recognised. This is not expected.
Can anybody suggest a method for echo cancellation
using Azure Speech SDK
in the chrome extension?
Azure speech SDK does not provide an option for noise cancellation as it uses the audio from the selected source (file or microphone) as is and passes it to the service.
A workaround could be to use an external microphone instead of the device microphone which can pickup other sounds or use the speech devices SDK which is actually a hardware device with the speech devices SDK and has the capability of noise suppression & echo cancellation.