10-701 Student Presentation - Combine cRBM and CNN for Object Recognition

Xinghai Hu, Minghao Ruan, Han Li We will present a novel deep network structure that combines the Convolutional Neural Network (CNN) and novel Convolutional Restricted Boltzmann Machine (cRBM). This combination can take advantage of the structural connections between stacked RBMs and Multi-layer Perceptrons (MLP). We will analyze the structure of both CNN and cRBM, and also analyze their similarities in structure. Besides, we will present with an essential component of a probabilistic max-pooling pro