Deep Convolutional Neural Networks for Image ...
The most common unsupervised schemes used contrastive divergence (Hinton, 2002) methods (see Lee, Grosse, Ranganath, & Ng, 2009), sparse constraints (Ranzato et al., 2006, 2007), or PSD (Kavukcuoglu et al., 2010; LeCun et al., 2010). In general, for these techniques, the feature extraction filters are trained such that representations at a ...