Tensorview: visualizing the training of convolutional neural network using paraview

Xinyu Chen; Qiang Guan; Xin Liang; Li-Ta Lo; Simon Su; Trilce Estrada; James Ahrens

doi:10.1145/3154842.3154846

Back

Conference proceeding

Tensorview: visualizing the training of convolutional neural network using paraview

Xinyu Chen, Qiang Guan, Xin Liang, Li-Ta Lo, Simon Su, Trilce Estrada and James Ahrens

Proceedings of the 1st Workshop on distributed infrastructures for deep learning, pp.11-16

DIDL '17

12/11/2017

DOI: https://doi.org/10.1145/3154842.3154846

Abstract and subjects

convolutional networks

paraview

visualization

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.

Metrics

1 Record Views

See more details

Details

Title: Tensorview
Authors/Creators - name: Xinyu Chen
Qiang Guan
Xin Liang
Li-Ta Lo
Simon Su
Trilce Estrada
James Ahrens
Publication Details: Proceedings of the 1st Workshop on distributed infrastructures for deep learning, pp.11-16
Series: DIDL '17
Publisher: ACM
Language: English
Resource Type: Conference proceeding
ISBN: 1450351697; 9781450351690
DOI: https://doi.org/10.1145/3154842.3154846