This project contains all the code for the experimental part of the paper "Adaptive recognition of machining features in sheet metal parts based on a graph class-incremental learning strategy."
The "Pre-train_Exp" folder contains files for the Sheet-metalNet's network architecture — "model.py"; the training process of Sheet-metalNet on the MFCAD++ dataset — "pretrain-mfcad.py"; and the training process of Sheet-metalNet on the SMCAD dataset — "pretrain-smcad.py".
The "Ablation-Comparison_Exp" folder contains all the Python code for the comparative experiments and ablation studies mentioned in the paper.
The "Inc-train_Exp" folder houses the code for all eight strategies of the incremental training experiments.
If you have any questions, please contact the first author of this paper, Ma Liuhuan, at ma_liuhuan@163.com.