Hyperspectral wafer surface classification based on a diffusion-model-enhanced incremental learning architecture

Abstract

Abstract Accurate wafer surface classification is essential for semiconductor process control, but its practical application is limited by the small number of labeled hyperspectral samples and the continuous emergence of previously unseen surface classes. To address these challenges, this study proposes a diffusion-model-enhanced incremental learning framework for hyperspectral wafer surface classification. A hyperspectral dataset containing eight representative wafer surface classes was constructed, covering normal material or structural states, process-related surface conditions, and potential defects. The diffusion model jointly learns the spatial structures and spectral correlations of hyperspectral data and generates supplementary samples for classifier training. During incremental updates, generated samples from previously learned classes are further used as rehearsal data to preserve historical knowledge, while the classification head is expanded to accommodate newly introduced classes. Experimental results show that diffusion-based augmentation improves the accuracy and Macro F1 score of the convolutional neural network by more than three percentage points compared with the non-augmented setting. In the continual-learning comparison, the proposed method achieves an accuracy of 95.98 percent and a Macro F1 score of 95.97 percent on the diffusion-enhanced dataset. The forgetting rate decreases from 6.79 percent to 4.64 percent, corresponding to a relative reduction of 31.66 percent. These results demonstrate that diffusion-generated rehearsal can improve hyperspectral wafer surface classification while effectively mitigating catastrophic forgetting during incremental updates.