Three projects in deep learning implemented with various results...
-
A neural network from scratch using numpy, as well as a generated binary image dataset containing four classes
-
A convolutional autoencoder and two convolution neural nets with identical architecture where one of the networks exploits the pretrained encoder from the autoencoder. Used for image recognition on four seperate datasets. Pytorch is used
-
Autoencoder and variational autoencoder implemented using tensorflow and keras