Deep Learning for Computer Vision

Autoencoders of raw images

An autoencoder is an unsupervised algorithm for generating efficient encodings. The input layer and the target output is typically the same. The layers between decrease and increase in the following fashion: 

The bottleneck layer is the middle layer with a reduced dimension. The left side of the bottleneck layer is called encoder and the right side is called decoder. An encoder typically reduces the dimension of the data and a decoder increases the dimensions. This combination of encoder and decoder is called an autoencoder. The whole network is trained with reconstruction error. Theoretically, the bottleneck layer ...