Generative Feature Replay with Orthogonal Weight Modification for Continual Learning

By Gehui Shen, Song Zhang, Xiang Chen, Zhihong Deng. 2020

25 citations recorded by the source corpus.

The ability of intelligent agents to learn and remember multiple tasks sequentially is crucial to achieving artificial general intelligence. Many continual learning (CL) methods have been proposed to overcome catastrophic forgetting which results from non i.i.d data in the sequential learning of neural networks. In this paper we focus on class incremental learning, a challenging CL scenario. For this scenario, generative replay is a promising strategy which generates and replays pseudo data for previous tasks to alleviate catastrophic forgetting. However, it is hard to train a generative model continually for relatively complex data. Based on recently proposed orthogonal weight modification (OWM) algorithm which can approximately keep previously learned feature invariant when learning new tasks, we propose to 1) replay penultimate layer feature with a generative model; 2) leverage a self-supervised auxiliary task to further enhance the stability of feature. Empirical results on several datasets show our method always achieves substantial improvement over powerful OWM while conventional generative replay always results in a negative effect. Meanwhile our method beats several strong baselines including one based on real data storage. In addition, we conduct experiments to study why our method is effective.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

1. Zihan Ye
2. Zhihong Deng
3. Fan Lyu
4. Huaping Liu
5. Song Zhang
6. Song Wang
7. Hao Liu
8. Shuai Wang

In Computer Vision and Pattern Recognition

1. Zihan Ye
2. Zhihong Deng
3. Fan Lyu
4. Huaping Liu
5. Song Zhang
6. Song Wang
7. Hao Liu
8. Shuai Wang

Citations in the corpus

• Continual Learning with Recursive Gradient Optimization (2022, background): “, 2021), or their generative models (Shin et al., 2017; Shen et al., 2020) to modify future training process.”
• Self-Supervised Learning Aided Class-Incremental Lifelong Learning (2020, background): “However, due to its own difficulty, generative replay has negative effect on OWM [40].”
• Multi-Domain Multi-Task Rehearsal for Lifelong Learning (2020, background): “…at the layers below the rehearsal layer and leaves the layers above free to learn at full pace; (3) by building generative model to synthesize data (Pseudo-rehearsal) (Shen et al. 2020; van de Ven and Tolias 2018; Lesort et al. 2019), the knowledge can be saved as parameters rather than data.”