Stacked What-Where Auto-encoders
This provides a novel approach for machine learning practitioners seeking unified models, though it appears incremental as it builds on existing convolutional and deconvolutional networks.
The paper tackles the problem of integrating discriminative and generative learning in a unified architecture, resulting in the stacked what-where auto-encoders (SWWAE) that supports supervised, semi-supervised, and unsupervised training without sampling.
We present a novel architecture, the "stacked what-where auto-encoders" (SWWAE), which integrates discriminative and generative pathways and provides a unified approach to supervised, semi-supervised and unsupervised learning without relying on sampling during training. An instantiation of SWWAE uses a convolutional net (Convnet) (LeCun et al. (1998)) to encode the input, and employs a deconvolutional net (Deconvnet) (Zeiler et al. (2010)) to produce the reconstruction. The objective function includes reconstruction terms that induce the hidden states in the Deconvnet to be similar to those of the Convnet. Each pooling layer produces two sets of variables: the "what" which are fed to the next layer, and its complementary variable "where" that are fed to the corresponding layer in the generative decoder.