Implementation of the LeNet-5 deep neural network model in Tensorflow. https://www.youtube.com/watch?v=ixkpBmKRZDw
Switch branches/tags
Nothing to show
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Failed to load latest commit information.
.gitignore lenet layers modified Jan 1, 2017
LeNet-Lab.ipynb test results printed Jan 1, 2017
README.md README title updated Apr 17, 2017
environment.yml lenet layers modified Jan 1, 2017
lenet.png lenet layers modified Jan 1, 2017

README.md

Implementation of the LeNet-5 deep neural network model.

LeNet-5 Architecture

Specs

Input Image is 28x28x1 and converted to 32x32x1 as per LeNet requirements.

Convolution layer 1: The output shape should be 28x28x6.

Activation 1: Your choice of activation function.

Pooling layer 1: The output shape should be 14x14x6.

Convolution layer 2: The output shape should be 10x10x16.

Activation 2: Your choice of activation function.

Pooling layer 2: The output shape should be 5x5x16.

Flatten layer: Flatten the output shape of the final pooling layer such that it's 1D instead of 3D.

Fully connected layer 1: This should have 120 outputs.

Activation 3: Your choice of activation function.

Fully connected layer 2: This should have 10 outputs.

Return the result of the 2nd fully connected layer from the LeNet function.