The goal of continual learning is to improve the performance of recognition models in learning sequentially arrived data. Although most existing works are established on the premise of learning from scratch, growing efforts have been devoted to incorporating the benefits of pre-training. However, how to adaptively exploit the pre-trained knowledge for each incremental task while maintaining its generalizability remains an open question. In this work, we present an extensive analysis for continual learning on a pre-trained model (CLPM), and attribute the key challenge to a progressive overfitting problem. Observing that selectively reducing the learning rate can almost resolve this issue in the representation layer, we propose a simple but extremely effective approach named Slow Learner with Classifier Alignment (SLCA), which further improves the classification layer by modeling the class-wise distributions and aligning the classification layers in a post-hoc fashion. Across a variety of scenarios, our proposal provides substantial improvements for CLPM (e.g., up to 49.76%, 50.05%, 44.69% and 40.16% on Split CIFAR-100, Split ImageNet-R, Split CUB-200 and Split Cars-196, respectively), and thus outperforms state-of-the-art approaches by a large margin. Based on such a strong baseline, critical factors and promising directions are analyzed in-depth to facilitate subsequent research.

翻译：持续学习的目标是提升识别模型在顺序学习数据时的性能。尽管现有研究多基于从零学习的假设，但越来越多的研究开始关注如何利用预训练的优势。然而，如何针对每个增量任务自适应地利用预训练知识同时保持其泛化性仍是一个开放性问题。本文对基于预训练模型的持续学习（CLPM）进行了深入分析，将关键挑战归因于渐进式过拟合问题。通过观察发现，在表征层选择性降低学习率几乎可以解决该问题，我们提出了一种名为"慢学习者与分类器对齐"（SLCA）的简单但极其有效的方法，该方法通过建模类级分布并以事后方式对齐分类层，进一步改进了分类层。在各种场景下，我们的方法为CLPM带来了显著提升（例如，在Split CIFAR-100、Split ImageNet-R、Split CUB-200和Split Cars-196上分别提升高达49.76%、50.05%、44.69%和40.16%），并以较大优势优于现有最先进方法。基于这一强基线，我们深入分析了关键因素和有前景的研究方向，以促进后续研究。