I'm currently working towards creating an app that implanents a CNN in a computer vision context. Ideally I want to train the model seperarely, and then be able to deploy it in OpenCV in Visual Studios (2017) for prototyping, and then in a mobile app as the actual product. What's the best way of achieving this?
Currently it seems like I have to use Tensorflow amd export the model to a server which I would call from Visual Studios or the app. Is there a better or easier way of doing this?
Thanks.
I would like to suggest you: http://www.easy-tensorflow.com/tf-tutorials/recurrent-neural-networks/bidirectional-rnn-for-classification
https://www.floydhub.com/
CNN is just an algorithm for a model, R, tensorflow...are framework that implement that model. In fact you can implement it in many languages but it's time consuming to write everything from scratch.
I think the easiest way is to use it as a service, meaning that your application only deals with collecting data. Send the data to server and then get back the result.
Use Python and link it with Visual Studio.
Follow this link:
https://docs.microsoft.com/en-us/visualstudio/python/tutorial-working-with-python-in-visual-studio-step-01-create-project?view=vs-2019
- Badges
- Science topic
- Similar topics
- Computer Science and Engineering
- Programming Languages