This work is proposed for Dual-Level Deep Evidential Fusion (DDEF): Integrating Multimodal Information for Enhanced Reliable Decision-Making in Deep Learning. DDEF is accurate, efficient, reliable, scalable, and very simple in design.
'utils.py' defined some shows details of Experiment 1.
'train.py' shows details of Experiment 1 on MNIST-SVHN.
You can also run the codes in Kaggle.
You can download SVHN data at http://ufldl.stanford.edu/housenumbers/
Other data can be found in /data
If you find this repository helpful, please consider citing:
@article{shao2024dual,
title={Dual-level Deep Evidential Fusion: Integrating multimodal information for enhanced reliable decision-making in deep learning},
author={Shao, Zhimin and Dou, Weibei and Pan, Yu},
journal={Information Fusion},
volume={103},
pages={102113},
year={2024},
publisher={Elsevier}
}