A Deep Generative Deconvolutional Image Model
This work addresses image classification and generation tasks for computer vision researchers, but it is incremental as it builds on existing deep learning frameworks.
The authors tackled image representation and analysis by developing a deep generative model using hierarchical convolutional dictionary learning and stochastic unpooling, achieving results highly competitive with similarly sized convolutional neural networks on benchmarks like ImageNet.
A deep generative model is developed for representation and analysis of images, based on a hierarchical convolutional dictionary-learning framework. Stochastic {\em unpooling} is employed to link consecutive layers in the model, yielding top-down image generation. A Bayesian support vector machine is linked to the top-layer features, yielding max-margin discrimination. Deep deconvolutional inference is employed when testing, to infer the latent features, and the top-layer features are connected with the max-margin classifier for discrimination tasks. The model is efficiently trained using a Monte Carlo expectation-maximization (MCEM) algorithm, with implementation on graphical processor units (GPUs) for efficient large-scale learning, and fast testing. Excellent results are obtained on several benchmark datasets, including ImageNet, demonstrating that the proposed model achieves results that are highly competitive with similarly sized convolutional neural networks.