Implementation-of-Hopfield-neural-network-for-Character-Recognition

Goal: Recognition of six first uppercase English letters by Hopfield neural network (even if they are noisy)

training set: the characters A, B, C, D, E, F, and Z

test set: noisy images

🚀 How to Run:

1- Open MATLAB.

2- Run the main GUI file:

hop

3- In the GUI:

I. Click "Create Network" to initialize the network with training patterns.

II. Enter a test character (e.g., A, B, ..., Z) in the textbox.

III. Click "Add Noise" to corrupt the input pattern.(It's Optional)

a = a + 0.1 * randn(size(a));

IV. Click "Run"

(This image is an example of the noisy given input “A”, the network attempts pattern recovery within a set number of steps, and the output at each step will be plotted in real-time, revealing the recovery process.)

License

You can view this project on open-source projects. You may fully use it for personal purposes, and it is licensed under the MIT License. You can share, alter, or modify it in any way you like.