How saccadic vision might help with the interpretability of deep networks

We describe how some problems (interpretability, lack of object-orientedness) of modern deep networks potentially could be solved by adapting a biologically plausible saccadic mechanism of perception. A sketch of such a saccadic vision model is proposed. Proof of concept experimental results are pro...

Full description

Saved in:
Bibliographic Details
Main Authors: Sereda, Iana, Osipov, Grigory
Format: Conference Proceeding
Language:English
Subjects:
Adaptation models
Biological system modeling
deep learning
ECG
interpretability by design
Neural networks
saccades
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:We describe how some problems (interpretability, lack of object-orientedness) of modern deep networks potentially could be solved by adapting a biologically plausible saccadic mechanism of perception. A sketch of such a saccadic vision model is proposed. Proof of concept experimental results are provided to support the proposed approach.
ISSN:2161-4407
DOI:10.1109/IJCNN52387.2021.9533289