I am new to the field of vision. To get a good understanding of the concepts, I wanted to look at source code of some CNN.

Can anybody suggest some place where I can find implementation of simple CNN like LeNet 5(preferred as I'm following it's tutorial) from starch in Matlab or python not using any library like tensorflow

Thanks in advance


Have a look at https://www.codeproject.com/Articles/16650/Neural-Network-for-Recognition-of-Handwritten-Digi. It contains a complete CNN for the recognition of handwritten digits which is built from scratch in C++. I dedicated about 3 full days to understand the source code. It was a challenging but very rewarding process which really helped me to get how CNNs work. The author also implements some advanced features such as a second order optimisation method (stochastic diagonal Levenberg-Marquardt) and threading to speed up the training process which is a big bonus in my opinion. It is often difficult to find practical examples of complex optimisation algorithms.

I would also recommend reading "Gradient-Based Learning Applied to Document Recognition" and "Efficient BackProp" by Y. LeCun et al.

You can also look at the assignments of the CS231n: Convolutional Neural Networks for Visual Recognition course.

| cite | improve this answer | |

Not the answer you're looking for? Browse other questions tagged or ask your own question.