Optimizing Reusable Knowledge for Continual Learning via Metalearning

By Julio Hurtado, Alain Raymond-Sáez, Álvaro Soto. 2021

46 citations recorded by the source corpus.

When learning tasks over time, artificial neural networks suffer from a problem known as Catastrophic Forgetting (CF). This happens when the weights of a network are overwritten during the training of a new task causing forgetting of old information. To address this issue, we propose MetA Reusable Knowledge or MARK, a new method that fosters weight reusability instead of overwriting when learning a new task. Specifically, MARK keeps a set of shared weights among tasks. We envision these shared weights as a common Knowledge Base (KB) that is not only used to learn new tasks, but also enriched with new knowledge as the model learns new tasks. Key components behind MARK are two-fold. On the one hand, a metalearning approach provides the key mechanism to incrementally enrich the KB with new knowledge and to foster weight reusability among tasks. On the other hand, a set of trainable masks provides the key mechanism to selectively choose from the KB relevant weights to solve each task. By using MARK, we achieve state of the art results in several popular benchmarks, surpassing the best performing methods in terms of average accuracy by over 10% on the 20-Split-MiniImageNet dataset, while achieving almost zero forgetfulness using 55% of the number of parameters. Furthermore, an ablation study provides evidence that, indeed, MARK is learning reusable knowledge that is selectively used by each task.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition, Meta & Continual Learning

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

In Computer Vision and Pattern Recognition

In Meta & Continual Learning

Citations in the corpus

A Comprehensive Survey of Continual Learning: Theory, Method and Application (2023, background): “MARK [186] maintains a set of shared weights that are incrementally updated with meta-learning and selectively masked to solve speciﬁc tasks.”
Anti-Retroactive Interference for Lifelong Learning (2022, background): “MARK [12] keeps a set of shared weights among tasks.”
Incremental Task Learning with Incremental Rank Updates (2022, background): “It has been a popular continual learning setup even in the very recent literature [7,14,11,36,31,40].”
CoSCL: Cooperation of Small Continual Learners is Stronger than a Big One (2022, neutral): “At the same time, we consider the SOTA methods that improve continual learning in terms of architecture, such as BNS [34] and MARK [17].”
Recent Advances of Continual Learning in Computer Vision: An Overview (2021, positive): “Hurtado et al. [182] proposed a method called Meta Reusable Knowledge (MARK), which has a single common knowledge base for all the learned tasks.”
How to Reuse and Compose Knowledge for a Lifetime of Tasks: A Survey on Continual Learning and Functional Composition (2022, methodology): “An additional algorithm instead avoids explicitly selecting tasks or layers to transfer, and directly meta-learns a set of features that maximize reuse when task-specific parameters mask the shared weights for transfer (Hurtado et al., 2021).”
Divide and not forget: Ensemble of selectively trained experts in Continual Learning (2024, methodology): “We compare SEED to: HAT (Serr`a et al., 2018), MARK (Hurtado et al., 2021), and BNS (Qin et al., 2021).”