Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Adding Italian model in pocketsphinx for offline recognition #101

Open
wants to merge 2 commits into
base: development
Choose a base branch
from
Open

Adding Italian model in pocketsphinx for offline recognition #101

wants to merge 2 commits into from

Conversation

geekypathak21
Copy link

Fixes fossasia/susi_linux#501
I added bash script to download Italian language model. Do you like this approach @stealthanthrax

Checklist

  • I have read the Contribution & Best practices Guide and my PR follows them.
  • My branch is up-to-date with the Upstream master branch.
  • The unit tests pass locally with my changes
  • I have added tests that prove my fix is effective or that my feature works
  • I have added necessary documentation (if appropriate)
  • All the functions created/modified in this PR contain relevant docstrings.

Test Passing

  • The SUSI Server must be building on the pi on bootup
  • The hotword detection should have a decent accuracy
  • SUSI Linux shouldn't crash when switching from online to offline and vice versa (failing as of now)
  • SUSI Linux should be able to boot offline when no internet connection available (failing)

Short description of what this resolves:

Changes proposed in this pull request:

@norbusan
Copy link
Member

Hi @himanshupathak21061998
How big are the model files all together necessary?

@geekypathak21
Copy link
Author

geekypathak21 commented Sep 26, 2019 •
edited

Hello @norbusan
The Italian model that I have just added will be of 9mb. If we add all languages it will take space of approx 1.5gb.

@sansyrox
Copy link
Member

@himanshupathak21061998 , which languages are occupying more space?
I believe we can support English, Italian, German, French, and Spanish as of now(some of the widely spoken languages).

@geekypathak21
Copy link
Author

Sorry for late reply.
I want to add more models so for which languages I should go for.
Size of languages is as follows:

  • US English is already installed by default
  • Italian :- 9.7mb
  • Spanish :- 57.1mb
  • German:- 159.4mb
  • French:- 46.1(small)mb
  • Spanish :- 57.1mb

@pranav1698
Copy link

@himanshupathak21061998 Your travis build is failing, can yo rectify that

@norbusan
Copy link
Member

norbusan commented Jan 7, 2020

@pranav1698 Failing travis builds is expected, since the patch adds an interactive question, which is a nogo. This has to be a command line option, otherwise we cannot use it in automatic builds

norbusan
norbusan reviewed Jan 7, 2020
View reviewed changes
install.sh
@@ -928,6 +928,24 @@ if [ $targetSystem = raspi ] ; then
sudo bash $INSTALLERDIR/raspi/access_point/wap.sh
fi

read -p "Do you want to download languages for offline recognition?[y/n] " lang
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Interaction is not possible, since we do build testing, as well as automated image creation. This needs to be a command line option or some combination.

@Orbiter
Copy link
Contributor

Orbiter commented Jan 31, 2020

to remove the interaction in the install script we have to decide if we want 300mb extra in the release for Pocket Sphinx. I doubt that we rely on Pocket Sphinx in the future so I believe that 300mb extra data should only be loaded in case that PocketSphinx is actually selected as tts.

@norbusan
Copy link
Member

Yeah, I doubt that this will be useful in the future. Maybe an option (script, button, webinterface?) to install direct onto the device.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers

@norbusan norbusan norbusan left review comments

Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.

More models for offline Speech Recognition
5 participants
@geekypathak21 @norbusan @sansyrox @pranav1698 @Orbiter