Visualization of Deep Convolutional Neural Networks

<p>Deep learning has achieved great accuracy in large scale image classification and scene recognition tasks, especially after the Convolutional Neural Network (CNN) model was introduced. Although a CNN often demonstrates very good classification results, it is usually unclear how or why a classification result is achieved. The objective of this thesis is to explore several existing visualization approaches which offer intuitive visual results. The thesis focuses on three visualization approaches: (1) image masking which highlights the region of image with high influence on the classification, (2) Taylor decomposition back-propagation which generates a per pixel heat map that describes each pixel's effect on the classification, and (3) Inception which generates a natural looking image based on the features maximizing the classification score. We explore two challenging visualization tasks, (1) visualizing a model classifying images based on the time when they are taken, and (2) visualizing a model of predicting plant phenotypes (specifically wheat heading percentage). The thesis demonstrates how these visualization approaches work for both the classification model and regression model, and evaluates the results on real-world imagery.</p>

Name thesis:degree_name

Master of Science (MS)

Level thesis:degree_level

Thesis

Discipline thesis:degree_discipline

Computer Science & Engineering

Year dc:date.available

2016

Author dc:creator

• Li, Dingwen

Contributors dc:contributor

• Robert Pless
• Yasu Furukawa Tao Ju

• Convolutional Neural Network
• Deep Learning
• Machine Learning
• Visualization
• Image Classification
• Engineering
• Other Computer Sciences

Statement dc:rights

• I have not registered my thesis with the U.S. Copyright Office, and do not intend to.

Language dc:language

English (en)

Identifier

https://openscholarship.wustl.edu/eng_etds/150

OAI identifier oai:identifier

oai:openscholarship.wustl.edu:eng_etds-1150