Attenuating Catastrophic Forgetting by Joint Contrastive and Incremental Learning

Quentin Ferdinand, Benoît Clement, Quentin Oliveau, Gilles Le Chenadec, Panagiotis Papadakis

CVPRW 2022 pp. 3781-3788

doi:10.1109/CVPRW56347.2022.00423 /cvprw/2022/ferdinand2022cvprw-attenuating/

Abstract

In class incremental learning, discriminative models are trained to classify images while adapting to new instances and classes incrementally. Training a model to adapt to new classes without total access to previous class data, however, leads to the known problem of catastrophic forgetting of the previously learnt classes. To alleviate this problem, we show how we can build upon recent progress on contrastive learning methods. In particular, we develop an incremental learning approach for deep neural networks operating both at classification and representation level which alleviates forgetting and learns more general features for data classification. Experiments performed on several datasets demonstrate the superiority of the proposed method with respect to well known state-of-the-art methods.

PDF CVPRW Semantic Scholar

Cite

Text

Ferdinand et al. "Attenuating Catastrophic Forgetting by Joint Contrastive and Incremental Learning." IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, 2022. doi:10.1109/CVPRW56347.2022.00423

Markdown

[Ferdinand et al. "Attenuating Catastrophic Forgetting by Joint Contrastive and Incremental Learning." IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, 2022.](https://mlanthology.org/cvprw/2022/ferdinand2022cvprw-attenuating/) doi:10.1109/CVPRW56347.2022.00423

BibTeX

@inproceedings{ferdinand2022cvprw-attenuating,
  title     = {{Attenuating Catastrophic Forgetting by Joint Contrastive and Incremental Learning}},
  author    = {Ferdinand, Quentin and Clement, Benoît and Oliveau, Quentin and Le Chenadec, Gilles and Papadakis, Panagiotis},
  booktitle = {IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops},
  year      = {2022},
  pages     = {3781-3788},
  doi       = {10.1109/CVPRW56347.2022.00423},
  url       = {https://mlanthology.org/cvprw/2022/ferdinand2022cvprw-attenuating/}
}