Class Incremental Learning (CIL) aims to sequentially learn new classes while
avoiding catastrophic forgetting of previous knowledge. We propose to use
Masked Autoencoders (MAEs) as efficient learners for CIL. MAEs were originally
designed to learn useful representations through reconstructive unsupervised
learning, and they can be easily integrated with a supervised loss for
classification. Moreover, MAEs can reliably reconstruct original input images
from randomly selected patches, which we use to store exemplars from past tasks
more efficiently for CIL. We also propose a bilateral MAE framework to learn
from image-level and embedding-level fusion, which produces better-quality
reconstructed images and more stable representations. Our experiments confirm
that our approach performs better than the state-of-the-art on CIFAR-100,
ImageNet-Subset, and ImageNet-Full. The code is available at
this https URL .

我们提出了使用 Masked Autoencoders (MAEs) 作为高效的学习器来解决 Class Incremental Learning (CIL) 的问题，并通过基于图像级和嵌入级融合的双边 MAE 框架获得更好质量的重建图像和更稳定的表示。实验证实了我们的方法在 CIFAR-100、ImageNet-Subset 和 ImageNet-Full 上优于现有技术。