mustafaghali/IndabaX-Rwanda-CNN-tutorial

# Deep Learning IndabaX Rwanda CNN tutorial * In this tutorial we will show how to design and train a ConvNet using PyTorch * We will also illustrate how the ConvNet makes use of specific assumptions * the ConvNet overview slides can be found on the tutorial Github repository ### When we use the CNNs we make use of the following assumptions about the input images signals * Many layers: compositionality * Convolutions: locality + stationarity of images * Pooling: Invariance of object class to translations