Publication
Title
Interpreting convolutional neural networks by explaining their predictions
Author
Abstract
We propose a method that exploits the feedback provided by visual explanation methods combined with pattern mining techniques to identify the relevant class-specific and class-shared internal units. In addition, we put forward a patch extraction approach to find faithfully class-specific and class-shared visual patterns. Contrary to the common practice in literature, our approach does not require pushing augmented visual patches through the model. Experiments on two CNN architectures show the effectiveness of the proposed method.
Language
English
Source (book)
2023 IEEE International Conference on Image Processing (ICIP), 08-11 October, 2023, Kuala Lumpur, Malaysia
Publication
IEEE , 2023
ISBN
978-1-7281-9835-4
DOI
10.1109/ICIP49359.2023.10222871
Volume/pages
p. 1685-1689
ISI
001106821001153
Full text (Publisher's DOI)
Full text (publisher's version - intranet only)
UAntwerpen
Faculty/Department
Research group
Project info
Multimodal Relational Interpretation for Deep Models.
Publication type
Subject
Affiliation
Publications with a UAntwerp address
External links
Web of Science
Record
Identifier
Creation 24.11.2023
Last edited 17.06.2024
To cite this reference