# Adversarial Feature Alignment: Avoid Catastrophic Forgetting in Incremental Task Lifelong Learning

arXiv: https://arxiv.org/abs/1910.10986 · HTML: https://mlsanity.com/paper/1910.10986/

## Key facts

- Cited by 4 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 24 Oct 2019.

## Built on by (most cited first, 4 of 4)

- [Towards Unsupervised Open World Semantic Segmentation](https://mlsanity.com/paper/2201.01073/) (2022)
- [Detecting and Learning the Unknown in Semantic Segmentation](https://mlsanity.com/paper/2202.08700/) (2022)
- [Deconfounded Lifelong Learning for Autonomous Driving via Dynamic Knowledge Spaces](https://mlsanity.com/paper/2603.14354/) (2026)
- [A Stepwise Distillation Learning Strategy for Non-differentiable Visual Programming Frameworks on Visual Reasoning Tasks](https://mlsanity.com/paper/2309.09809/) (2023)

## Abstract (arXiv)

Human beings are able to master a variety of knowledge and skills with ongoing learning. By contrast, dramatic performance degradation is observed when new tasks are added to an existing neural network model. This phenomenon, termed as Catastrophic Forgetting, is one of the major roadblocks that prevent deep neural networks from achieving human-level artificial intelligence. Several research efforts, e.g. Lifelong or Continual learning algorithms, have been proposed to tackle this problem. However, they either suffer from an accumulating drop in performance as the task sequence grows longer, or require to store an excessive amount of model parameters for historical memory, or cannot obtain competitive performance on the new tasks. In this paper, we focus on the incremental multi-task image classification scenario. Inspired by the learning process of human students, where they usually decompose complex tasks into easier goals, we propose an adversarial feature alignment method to avoid catastrophic forgetting. In our design, both the low-level visual features and high-level semantic features serve as soft targets and guide the training process in multiple stages, which provide sufficient supervised information of the old tasks and help to reduce forgetting. Due to the knowledge distillation and regularization phenomenons, the proposed method gains even better performance than finetuning on the new tasks, which makes it stand out from other methods. Extensive experiments in several typical lifelong learning scenarios demonstrate that our method outperforms the state-of-the-art methods in both accuracies on new tasks and performance preservation on old tasks.