User Feedback on 'Implementation of Google Teachable Machine' [GSOC'22]

Hello :wave: everyone,

I am Yugal Agarwal, a Google Summer of Code 2022 contributor at MIT App Inventor. I will be working on the project "Implementation of Teachable Machine as an Extension" under the mentorship of David Kim and Evan Patton(@ewpatton). This post is made in order to get the user feedback.

Project Description

This project aims to create an extension to integrate Google Teachable Machine to app inventor. With this we would be able to directly integrate the models that we create in teachable machine to the app inventor.


Through this project we don't need to have different extension for different types of models like audio, image classificatin and pose model will have a single extension. And using this extension will be similar how we would use Personal Image classfication extension. We just have to export our model from teachable machine and we would be able to use it in similar way we can use the model in PIC extension.

Personal Image Classification is used for image classification. Below is the link to one of the apps made with PIC extension Personal Image Classifier: PICaboo.

As for Google Teachable Machine, it is a web-based tool that makes creation of machine learning models fast, easy and with no prior any coding knowledge. For now it provides 3 kinds of model image classification, audio classification and Pose classification.

Yugal Agarwal.


Good luck with your project. :+1::+1::+1:

Hi Yugal, good luck with your project! I hope the outcome will be as beautiful as your user name. And don't upset your mentor by misspelling his name :unamused:
What I would like to see is more information, for example a small description of the Google Teachable Machine or at least a pointer to where I can find information.
Maybe you can also tell about your deliverables, so that also people who have no clue (like me) what the Personal Image Classification Extension is and what the similarity is.
In this way you will get a larger audience and probably more feedback.

Thanks for the complement :grin:. And I am sorry for misspelling the name :sweat_smile:. And is there any way to edit the post or do i have to create new post for the same?

If you click on the three dots at the bottom of your post you get this menu

Click on the pencil to edit your post.

I am only getting this.
Not the edit[pencil] one.

Do you see it now? I increased your Trustlevel since you are a GSOC participant.

1 Like

Yes now its visible. Thanks

1 Like

How's it progressing?

Very interested in this! As I find TM to be one of the most straightforward tools to create models. Hope all is going well :slight_smile:

Any news, so I'd love to be able to use this in my projects!

Otherwise I'm going to have to learn to use android studio instead, as the MIT classifier dosen't work with my dataset, as it's too big!

Yeah, I am working on it. Trying to complete it as soon as possible.

Could you put your work so far on github, with a TODO list?

For now the extension works for image classification models only, work on audio and pose is still in progress.
Project Link

1 Like

Feedback is appreciated, whether there is something that needs to be improved, or anything else that you think could be added. Thanks :grinning_face_with_smiling_eyes:

1 Like

Thanks a lot for your hard work! I will try and use it as soon as I have some time and let you know what happened.