Online Fast Adaptation and Knowledge Accumulation (OSAKA): a New Approach to Continual Learning

By Massimo Caccia, Pau Rodr'iguez L'opez, Oleksiy Ostapenko, Fabrice Normandin, Min Lin, Lucas Page-Caccia, Issam Hadj Laradji, Irina Rish, Alexandre Lacoste, David Vázquez and 1 more. 2020

80 citations recorded by the source corpus.

Subfields: 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 Meta & Continual Learning

1. Massimo Caccia
2. Xingxing Zhang
3. Irina Rish
4. Liyuan Wang
5. Hang Su
6. Timothée Lesort
7. Jun Zhu
8. Dacheng Tao

Citations in the corpus

• A Comprehensive Survey of Continual Learning: Theory, Method and Application (2023, background): “OSAKA[43]proposes ahybridobjectiveofknowledgeaccumulationandfastadap-tation,whichcanberesolvedbyobtainingagoodinitializa-tionwithmeta-trainingandthenincorporatingknowledgeof incrementaltasksintotheinitialization.”
• An Empirical Investigation of the Role of Pre-training in Lifelong Learning (2021, background): “Caccia et al. (2020) pro-pose a two-phase continual learning scenario where the first phase is pre-training (using MAML (Finn et al., 2017)) and the second phase involves continual deployment with task revisiting.”
• Optimizing Reusable Knowledge for Continual Learning via Metalearning (2021, methodology): “It has been used as a training strategy in sequence learning [43, 44] and continual learning scenarios [45, 46, 47, 48].”
• Same State, Different Task: Continual Reinforcement Learning without Interference (2021, background): “Meta-learning pre-training has been explored to augment continual learning (and contextdependent targets can allow interference but is not explored) (Caccia et al. 2020).”
• Learn Continually, Generalize Rapidly: Lifelong Knowledge Accumulation for Few-shot Learning (2021, background): “Caccia et al. (2020) proposed a setup where models visit a sequence of potentially re-occuring tasks and measured online cumulative performance as metrics.”
• Towards Continual Reinforcement Learning: A Review and Perspectives (2020, background): “Indeed, this phased learning setting can even be adopted for continual learning deployment at meta-test time by training the inner loop continually over a sequence of tasks (Javed and White, 2019; Spigler, 2019; Beaulieu et al., 2020; Caccia et al., 2020; Kim et al., 2020).”
• MgSvF: Multi-Grained Slow versus Fast Framework for Few-Shot Class-Incremental Learning (2020, neutral): “Recently, there has been a large body of research in incremental learning [18], [19], [20], [21], [22], [23], [24], [25], [26], [27], [28], [29], [30], [31], [32], [33], [34], [35], [36], [37], [38], [39], [40], [41], [42], [43].”
• Knowledge Distillation: A Survey (2020, background): “Recently, an increasing number of KD variants, which are based on lifelong learning, have been developed (Jang et al. 2019; Flennerhag et al. 2019; Peng et al. 2019b; Liu et al. 2019e; Lee et al. 2019b; Zhai et al. 2019; Zhou et al. 2020; Shmelkov et al. 2017; Li and Hoiem 2017; Caccia et al. 2020).”
• A Policy Gradient Algorithm for Learning to Learn in Multiagent Reinforcement Learning (2020, methodology): “…for a dynamic Markov chain of other agent policies, we are leveraging an approach that has recently become popular for supervised learning called meta-continual learning (Riemer et al., 2019; Javed & White, 2019; Spigler, 2019; Beaulieu et al., 2020; Caccia et al., 2020; Gupta et al., 2020).”
• Update Frequently, Update Fast: Retraining Semantic Parsing Systems in a Fraction of Time (2021, background): “A usual task-incremental setup (Caccia et al., 2020; de Masson d’Autume et al., 2019) is far from the real-world applications.”
• Sequoia: A Software Framework to Unify Continual Learning Research (2021, methodology): “Some cases where this setting has been studied are [12, 52] for RL and [7, 23, 22] for SL.”
• Half-Real Half-Fake Distillation for Class-Incremental Semantic Segmentation (2021, unknown): “The problem has been extensively studied for image classification tasks [9, 3, 47].”
• Learning where to learn: Gradient sparsity in meta and continual learning (2021, methodology): “2 We finally consider another online learning setting in which the underlying task is concealed from the learner and can randomly change at each step, potentially going back to previously seen tasks [48, 18, 7].”
• CLEVA-Compass: A Continual Learning EValuation Assessment Compass to Promote Research Transparency and Comparability (2021, methodology): “For instance, one could regularize parameter deviations over time according to a supervised importance measure, assume a supervised representation pre-training step as in OSAKA, or rely on various unsupervised quantities, such as exponential moving averages.”