German speech input ends input with superfluous words

[bytinbit]

I have been using nerd-dictation for a while and it's fantastic - open-source, adaptable, hackable, Python, Linux-friendly :) Thanks!

But there's a strange thing happening: English input works without any problems, German input however nearly always prints nein (no) or einen (one, pronoun) at the end of a spoken information chunk. I have no idea why.

System settings: Linux Mint Cinnamon 20.2 (Ubuntu 20.04.1)

I invoke both via keyboard shortcuts that call a bash script. Commands in the bash script (I skipped the venv-related paths etc)

# German
nerd-dictation begin --vosk-model-dir ~/opt/nerd-dictation/model-de --numbers-as-digits --timeout 5 --punctuate-from-previous-timeout 3

# English
nerd-dictation begin --vosk-model-dir ~/opt/nerd-dictation/model-en-us --numbers-as-digits --timeout 5 --punctuate-from-previous-timeout 3

Models are the full versions: German is vosk-model-de-0.21.zip, English is vosk-model-en-us-0.22.zip.

I suppose it might be related to:

• xdotool or something like that
• the applications in which I use nerd-dictation
• ... something else?

I am at a loss how to debug it or discover the origin of the superfluous words. Any ideas, explanations, possibilities?

[iradraconis]

same problem here

[iradraconis]

may I also ask how you deal with capitalizing german nouns? I found a solution, that works pretty well, but it is rather slow.

[bytinbit]

Simple answer: I don't deal with it at all... I use it mainly for chats, which are close to spoken speech and in which correct capitalization is more or less irrelevant. I recently tried a setup with KDE-Connect on my phone + speech-to-text + linked to my Linux host, since Android's speech-to-text recognizes German capitalization rules, but sadly it does not work.

[bytinbit]

@iradraconis do you also have the same superfluous words einen and nein or are they different?

[iradraconis]

@iradraconis do you also have the same superfluous words einen and nein or are they different?

exactly these words. More often "einen".

[papoteur-mga]

Hello, The most probable is that problem comes from the vosk model for german itself. In French, I see frequently "hum" without real reason. xdotool takes characters and sends them as a key simulated. It doesn't introduce any word.

[2No2Name]

This happens for me when the microphone is muted (muted both physically and in the system sound preferences).

[Lunatik00]

i think the problem might be the same that the one i had in spanish, so i have a workaround posted in #43