SC17 Denver, CO

P53: TensorViz: Visualizing the Training of Convolutional Neural Network Using ParaView

Authors: Xinyu Chen (University of New Mexico), Qiang Guan (Los Alamos National Laboratory), Xin Liang (University of California, Riverside), Li-Ta Lo (Los Alamos National Laboratory), Simon Su (US Army Research Laboratory), Trilce Estrada (University of New Mexico), James Ahrens (Los Alamos National Laboratory)

Abstract: Deep Convolutional Networks have been very successful in visual recognition tasks recently. Previous works visualize learned features at different layers to help people understand how CNNs learn visual recognition tasks. However, they do not help to accelerate the training process. We use ParaView to provides both qualitative and quantitative visualization that help understand the learning procedure, tune the learning parameters, and direct merging and pruning of neural networks.
Award: Best Poster Finalist (BP): no

Poster: pdf
Two-page extended abstract: pdf

Poster Index