Explain details about the files

[Shakib-IO]

Can you describe the model's files I mean what files put where and how it's linked and how the overall path of files

[MikeM711]

Sure, all of the models' files reside inside of data/model-development. For the model, I just ran final_model_1.py x3 times and final_model_2.py x2 times for 5 total .h5 models. That's it when it comes to the deep learning part.

After that, I imported these model files into my Django web backend project inside handwriting/utils/models/. In the backend, I did some manipulation with incoming images that come from the frontend, and made predictions using these .h5 files.

I chose React as the web frontend to communicate these images to Django. The React frontend communicates with the backend using blobs (which were once images of the canvas). Eventually I get a response from the backend on what it predicts and display the result on the frontend.

There is a lot of web development involved in this project, which is more of my strong suit. The actual AI part is just a small section of the project (just the x5 .h5 model files that are used for prediction).

[Shakib-IO]

Thanks for reply But can give me some guide where I can start this project for my own purpose .

…

On Tue, Nov 24, 2020 at 2:24 AM Michael McCabe ***@***.***> wrote: Sure, all of the models' files reside inside of data/model-development. For the model, I just ran final_model_1.py x3 times and final_model_2.py x2 times for 5 total models. That's it when it comes to the deep learning part. After that, I imported these model files into my Django web backend project. In the backend, I did some manipulation with incoming images that come from the frontend, and made predictions using these .h5 files. I chose React as the web frontend to communicate these images to Django. The React frontend communicates with the backend using blobs (which were once images of the canvas). Eventually I get a response from the backend on what it predicts and display the result on the frontend. There is a lot of web development involved in this project, which is more of my strong suit. The actual AI part (with models and everything) is just a small section of the project. — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#3 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/APSXPBQ2RGBP2FY3IKNAA4LSRLAH5ANCNFSM4T7D77MQ> .

[MikeM711]

I send the canvas image blob over to Django inside the /src/components/Canvas/Canvas.js React component and I receive the blob at the corresponding Django POST endpoint: which is in /handwriting/views.py

That might help you out for where the connection is to start you off