ViT-MUL: A Baseline Study on Recent Machine Unlearning Methods Applied to Vision Transformers
This work tackles the problem of machine unlearning for Vision Transformers, which is incremental as it extends existing methods to a new model architecture.
The paper conducted a baseline study applying recent machine unlearning methods to Vision Transformers, providing comprehensive experiments and findings to address the lack of research in this area.
Machine unlearning (MUL) is an arising field in machine learning that seeks to erase the learned information of specific training data points from a trained model. Despite the recent active research in MUL within computer vision, the majority of work has focused on ResNet-based models. Given that Vision Transformers (ViT) have become the predominant model architecture, a detailed study of MUL specifically tailored to ViT is essential. In this paper, we present comprehensive experiments on ViTs using recent MUL algorithms and datasets. We anticipate that our experiments, ablation studies, and findings could provide valuable insights and inspire further research in this field.