.. _paper-xder:

Class-Incremental Continual Learning into the eXtended DER-verse
================================================================

    - :fa:`circle-check` `10.1109/TPAMI.2022.3206549 <https://doi.org/10.1109/TPAMI.2022.3206549>`_
    - :fa:`calendar` September 2022
    - :fa:`scroll` `IEEE TPAMI <https://ieeexplore.ieee.org/xpl/RecentIssue.jsp?punumber=34>`_
    - :fa:`tags` :bdg-primary:`Continual Learning` :bdg-primary:`Distillation` :bdg-primary:`Rehearsal` :bdg-primary:`Class-Incremental`

The study focuses on Class-Incremental Continual Learning, aiming to address the challenge of catastrophic forgetting in Deep Networks. The researchers improve upon Dark Experience Replay (DER), with a new method called **eXtended-DER** (**X-DER**). X-DER revises its replay memory to incorporate new information and facilitate learning of previously unseen classes. Results show that X-DER outperforms existing methods on standard benchmarks like CIFAR-100 and miniImageNet, as well as a new benchmark introduced in the study. Ablation studies highlight the importance of Knowledge Distillation and the role of flatter minima in continual learning setups.