[TPAMI 2025] Revisiting Essential and Nonessential Settings of Evidential Deep Learning

Authors: Mengyuan Chen, Junyu Gao, Changsheng Xu.

Affiliations: Institute of Automation, Chinese Academy of Sciences

Paper: https://arxiv.org/pdf/2410.00393

Dependencies:

Here we list our used requirements and dependencies. Theoretically, the specific versions of dependencies should not affect the performance of the method.

• GPU: GeForce RTX 3090
• Python: 3.8.5
• PyTorch: 1.12.0
• Pandas: 1.1.3
• Scikit-learn: 1.0.1
• Wandb: 0.12.6
• Tqdm: 4.62.3

Data preparation:

The required datasets (CIFAR-10/CIFAR-100/SVHN/GTSRB/LFWPeople/Places365/Food101) will be automatically downloaded if your server has an Internet connection.

Pre-trained models:

The pre-trained models of EDL, I-EDL, R-EDL, and Re-EDL can be downloaded from Google Disk.

They need to be unzipped and put in the directory './code_classical/saved_models/'.

Test pre-trained models:

Just run:

python main.py --configid "2_cifar10/cifar10-{method-name}-test" --suffix test

where "method-name" should be replaced by "edl", "iedl", "redl", or "reedl".

Train from scratch:

Just run:

python main.py --configid "2_cifar10/cifar10-{method-name}-train" --suffix test