Continuous repetition of "not able to recognize" error messages

[pranavkarthik10]

After installation of susi_installer on Ubuntu, I configured and ran susi_linux using python3. Right when it started even though I had made no noise, it began a continuous repetition of 'Sorry, I couldn't recognize' messages were said immediately before stopping after a few seconds.

I also found that the sensitivity of the speech-to-text system could be minimized because of it catching up on the slightest sounds nearby. A possible starting phrase such as 'Hey Susi' or just 'SUSI' before stating any message from the user's end could be used to not activate the assistant unnecessarily.

[norbusan]

We unfortunately don't have control over the microphone settings on the desktop for now. You need to adjust the correct microphone and the recording level manually.

Can you please stop susi-python process, and start it by hand with

cd ~/SUSI.AI/susi_linux/
python3 -m main -v -v

and attach the output here? Thanks

[pranavkarthik10]

Attached output here: https://pastebin.com/cSe2ZyyT

[norbusan]

A few things:

• you don't have the susi server running, right? susi_linux says local server is down in the first line
• what is contained in ~/SUSI.AI/config.json

Then please re-test with the following commands:

cd ~/SUSI.AI
PYTHONPATH=/home/norbert/SUSI.AI/susi_linux:$PYTHONPATH
export PYTHONPATH
python3 -m main -v -v

But yeah, I see some problems here myself, too ... I see that

[pranavkarthik10]

Unfortunately my computer in which I was running the VM for linux is currently not working well. Sorry for any delays.

[norbusan]

I have fixed these problems in the susi_linux code, thus closing the issue