Code for paper "Understanding the Effect of Different Priori Knowledge on CNN Fault Interpreter"
| Filename | usage |
|---|---|
| figures*.ipynb | They are used for generating figures in the above paper |
| model_zoo | benchmark models |
| cococlasses.ipynb | It checks the classes involved in coco dataset and class distribution. |
| savePredNpy_thebetestall-norm-final.ipynb | It is used for generating fault predictions |
| trainPretrainedModelFinal.py | It can used for pretraining and fine-tuning. |