TY  - CONF
AU  - Kai Wang
AU  - Xialei Liu
AU  - Andrew Bagdanov
AU  - Luis Herranz
AU  - Shangling Jui
AU  - Joost Van de Weijer
A2  - CVPRW
PY  - 2022//
TI  - Incremental Meta-Learning via Episodic Replay Distillation for Few-Shot Image Recognition
BT  - CVPR 2022 Workshop on Continual Learning (CLVision, 3rd Edition)
SP  - 3728
EP  - 3738
KW  - Training
KW  - Computer vision
KW  - Image recognition
KW  - Upper bound
KW  - Conferences
KW  - Pattern recognition
KW  - Task analysis
N2  - In this paper we consider the problem of incremental meta-learning in which classes are presented incrementally in discrete tasks. We propose Episodic Replay Distillation (ERD), that mixes classes from the current task with exemplars from previous tasks when sampling episodes for meta-learning. To allow the training to benefit from a large as possible variety of classes, which leads to more gener-alizable feature representations, we propose the cross-task meta loss. Furthermore, we propose episodic replay distillation that also exploits exemplars for improved knowledge distillation. Experiments on four datasets demonstrate that ERD surpasses the state-of-the-art. In particular, on the more challenging one-shot, long task sequence scenarios, we reduce the gap between Incremental Meta-Learning andthe joint-training upper bound from 3.5% / 10.1% / 13.4% / 11.7% with the current state-of-the-art to 2.6% / 2.9% / 5.0% / 0.2% with our method on Tiered-ImageNet / Mini-ImageNet / CIFAR100 / CUB, respectively.
L1  - http://158.109.8.37/files/WLB2022.pdf
UR  - http://dx.doi.org/10.1109/CVPRW56347.2022.00417
N1  - LAMP; 600.147
ID  - Kai Wang2022
ER  -