掩码自编码器是高效的类别增量学习器
Masked Autoencoders are Efficient Class Incremental Learners
AI总结：
针对类别增量学习的灾难性遗忘问题，提出将掩码自编码器作为高效学习器，利用其重构能力高效存储过往样本，并提出双边MAE框架实现图像级与嵌入级融合，在多个基准数据集上取得优于现有最优方法的性能。
摘要：
类别增量学习（Class Incremental Learning, CIL）旨在依次学习新类别，同时避免对已有知识的灾难性遗忘。我们提出将掩码自编码器（Masked Autoencoders, MAEs）用作CIL的高效学习器。MAEs最初设计用于通过重构式无监督学习学习有用的表征，且可轻松与监督损失结合以实现分类。此外，MAEs能从随机选取的图像块中可靠重构原始输入图像，我们利用该特性为CIL更高效地存储过往任务的样本。我们还提出一种双边MAE框架，从图像级和嵌入级融合中学习，可生成质量更高的重构图像和更稳定的表征。我们的实验证实，该方法在CIFAR-100、ImageNet-Subset和ImageNet-Full上的表现优于现有最优方法。代码可在https://github.com/scok30/MAE-CIL 获取。
英文摘要：
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 https://github.com/scok30/MAE-CIL .
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
