English

Continual Learning in Open-vocabulary Classification with Complementary Memory Systems

Computer Vision and Pattern Recognition 2024-10-08 v3

Abstract

We introduce a method for flexible and efficient continual learning in open-vocabulary image classification, drawing inspiration from the complementary learning systems observed in human cognition. Specifically, we propose to combine predictions from a CLIP zero-shot model and the exemplar-based model, using the zero-shot estimated probability that a sample's class is within the exemplar classes. We also propose a "tree probe" method, an adaption of lazy learning principles, which enables fast learning from new examples with competitive accuracy to batch-trained linear models. We test in data incremental, class incremental, and task incremental settings, as well as ability to perform flexible inference on varying subsets of zero-shot and learned categories. Our proposed method achieves a good balance of learning speed, target task effectiveness, and zero-shot effectiveness. Code will be available at https://github.com/jessemelpolio/TreeProbe.

Keywords

Cite

@article{arxiv.2307.01430,
  title  = {Continual Learning in Open-vocabulary Classification with Complementary Memory Systems},
  author = {Zhen Zhu and Weijie Lyu and Yao Xiao and Derek Hoiem},
  journal= {arXiv preprint arXiv:2307.01430},
  year   = {2024}
}

Comments

Accepted by Transactions on Machine Learning Research (TMLR)

R2 v1 2026-06-28T11:21:24.215Z