Skip to content
This repository has been archived by the owner on Oct 9, 2022. It is now read-only.

First Windows Form Application to Recognize MNIST Handwritten Digits using TensorflowSharp for C#. Using CNN Classifier.

License

Notifications You must be signed in to change notification settings

captain-pool/TensorflowSharp-MNIST

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

9 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

First Windows Form Application to Recognize MNIST Handwritten Digits using TensorflowSharp for C#. Using CNN Classifier.

Usage: (To Retrain model)

  • cd python/
    python3 graph.py
    
  • Freeze Graph for usage:
        freeze_graph --input_graph=tmp/model/tf_graph.pb \
        --input_checkpoint=tmp/model/weights.ckpt --input_binary=True --output_node_names=output \
        --output_graph=output.pb

Build and Run! The Compiled .exe can be found in bin/debug/TF Test.exe

Don't move the compiled binary as it depends on the ouput.pb file fixed for the current folder hierarchy

Download the mnist handwriting images(28x28) from Here, unzip it, and use it to run the model. else feed any 28x28 handwritten image, and it should work.

Screenshot.png

About

First Windows Form Application to Recognize MNIST Handwritten Digits using TensorflowSharp for C#. Using CNN Classifier.

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published