TensorView: Visualizing the Training of Convolutional Neural Network using Paraview
Abstract
Convolutional Neural Networks(CNNs) have been widely used in visual recognition tasks recently. Previous works visualize learning features at different layers to help people to understand how CNNs learn visual recognition tasks. However they only provide qualitative description and do not help to accelerate the training process. We present TensorView to enable Paraview to visualize the evolution of CNNs. TensorView provides both qualitative and quantitative visualization that help understand the learning procedure, tune the learning parameters, direct merging and pruning of neural networks.
Recommended Citation
X. Chen et al., "TensorView: Visualizing the Training of Convolutional Neural Network using Paraview," Proceedings of the 1st Workshop on Distributed Infrastructures for Deep Learning (2017, Las Vegas, NV), pp. 11 - 16, Association for Computing Machinery (ACM), Dec 2017.
The definitive version is available at https://doi.org/10.1145/3154842.3154846
Meeting Name
1st Workshop on Distributed Infrastructures for Deep Learning, DIDL 2017, Part of Middleware '17 (2017: Dec. 11-15, Las Vegas, NV)
Department(s)
Computer Science
Keywords and Phrases
Convolutional Networks; Paraview; Visualization
International Standard Book Number (ISBN)
978-145035169-0
Document Type
Article - Conference proceedings
Document Version
Citation
File Type
text
Language(s)
English
Rights
© 2017 Association for Computing Machinery (ACM), All rights reserved.
Publication Date
11 Dec 2017
