Learning Invariant Representation for Continual Learning

Ghada A.Z.N. Sokar, Decebal Constantin Mocanu, Mykola Pechenizkiy

Research output: Chapter in Book/Report/Conference proceedingConference contributionAcademicpeer-review

Abstract

Continual learning aims to provide intelligent agents that are capable of learning continually a sequence of tasks, building on previously learned knowledge. A key challenge in this learning paradigm is catastrophically forgetting previously learned tasks when the agent faces a new one. Current rehearsal-based methods show their success in mitigating the catastrophic forgetting problem by replaying samples from previous tasks during learning a new one. However, these methods are infeasible when the data of previous tasks is not accessible. In this work, we propose a new pseudo-rehearsal based method, named learning Invariant Representation for Continual Learning (IRCL), in which class-invariant representation is disentangled from a conditional generative model and jointly used with class-specific representation to learn the sequential tasks. Disentangling the shared invariant representation helps to learn continually a sequence of tasks, while being more robust to forgetting and having better knowledge transfer. We focus on class incremental learning where there is no knowledge about task identity during inference.We empirically evaluate our proposed methods on two well-known benchmarks for continual learning: split MNIST and split Fashion MNIST. The experimental results show that our proposed method outperforms regularization-based methods by a big margin and is better than the state-of-the-art pseudo-rehearsal-based methods. Finally, we analyze the role of the shared invariant representation in mitigating the forgetting problem especially when the number of replayed samples for each previous task is small.
Original languageEnglish
Title of host publicationMeta-Learning for Computer Vision (MeL4CV) workshop at AAAI Conference on Artificial Intelligence (AAAI-21)
Publication statusAccepted/In press - 21 Dec 2020
Event35th Conference on Artificial Intelligence, AAAI 2021 - Virtual Conference
Duration: 2 Feb 20219 Feb 2021
Conference number: 35
https://aaai.org/Conferences/AAAI-21/

Conference

Conference35th Conference on Artificial Intelligence, AAAI 2021
Abbreviated titleAAAI 2021
Period2/02/219/02/21
Internet address

Fingerprint Dive into the research topics of 'Learning Invariant Representation for Continual Learning'. Together they form a unique fingerprint.

Cite this