A Dashboard for Interactive Convolutional Neural Network Training And Validation Through Saliency Maps

Loading...
Thumbnail Image
Date
2023
Journal Title
Journal ISSN
Volume Title
Publisher
The Eurographics Association
Abstract
Quali-quantitative methods provide ways for interrogating Convolutional Neural Networks (CNN). For it, we propose a dashboard using a quali-quantitative method based on quantitative metrics and saliency maps. By those means, a user can discover patterns during the training of a CNN. With this, they can adapt the training hyperparameters of the model, obtaining a CNN that learned patterns desired by the user. Furthermore, they neglect CNNs which learned undesirable patterns. This improves users' agency over the model training process.
Description

CCS Concepts: Computing methodologies -> Artificial intelligence

        
@inproceedings{
10.2312:evp.20231054
, booktitle = {
EuroVis 2023 - Posters
}, editor = {
Gillmann, Christina
and
Krone, Michael
and
Lenti, Simone
}, title = {{
A Dashboard for Interactive Convolutional Neural Network Training And Validation Through Saliency Maps
}}, author = {
Cech, Tim
and
Simsek, Furkan
and
Scheibel, Willy
and
Döllner, Jürgen
}, year = {
2023
}, publisher = {
The Eurographics Association
}, ISBN = {
978-3-03868-220-2
}, DOI = {
10.2312/evp.20231054
} }
Citation