1 paper
Yichen Wen, Zhiquan Tan, Kaipeng Zheng +2
Continual learning requires learning incremental tasks with dynamic data distributions. So far, it has been observed that employing a combination of contrastive loss and distillati…