Dynamic Self-Supervised Teacher-Student Network Learning

TY - JOUR

T1 - Dynamic Self-Supervised Teacher-Student Network Learning

AU - Ye, Fei

AU - Bors, Adrian Gheorghe

N1 - © 2022 IEEE This is an author-produced version of the published paper. Uploaded in accordance with the publisher’s self-archiving policy. Further copying may not be permitted; contact the publisher for details

PY - 2022/11/10

Y1 - 2022/11/10

N2 - Lifelong learning (LLL) represents the ability of an artificial intelligence system to learn successively a sequence of different databases. In this paper we introduce the Dynamic Self-Supervised Teacher-Student Network (D-TS), representing a more general LLL framework, where the Teacher is implemented as a dynamically expanding mixture model which automatically increases its capacity to deal with a growing number of tasks. We propose the Knowledge Discrepancy Score (KDS) criterion for measuring the relevance of the incoming information characterizing a new task when compared to the existing knowledge accumulated by the Teacher module from its previous training. The KDS ensures a light Teacher architecture while also enabling to reuse the learned knowledge whenever appropriate, accelerating the learning of given tasks. The Student module is implemented as a lightweight probabilistic generative model. We introduce a novel self-supervised learning for the Student that allows to capture cross-domain latent representations from the entire knowledge accumulated by the Teacher as well as from novel data. We perform several experiments which show that D-TS can achieve the state of the art results in LLL while requiring fewer parameters than other methods.

AB - Lifelong learning (LLL) represents the ability of an artificial intelligence system to learn successively a sequence of different databases. In this paper we introduce the Dynamic Self-Supervised Teacher-Student Network (D-TS), representing a more general LLL framework, where the Teacher is implemented as a dynamically expanding mixture model which automatically increases its capacity to deal with a growing number of tasks. We propose the Knowledge Discrepancy Score (KDS) criterion for measuring the relevance of the incoming information characterizing a new task when compared to the existing knowledge accumulated by the Teacher module from its previous training. The KDS ensures a light Teacher architecture while also enabling to reuse the learned knowledge whenever appropriate, accelerating the learning of given tasks. The Student module is implemented as a lightweight probabilistic generative model. We introduce a novel self-supervised learning for the Student that allows to capture cross-domain latent representations from the entire knowledge accumulated by the Teacher as well as from novel data. We perform several experiments which show that D-TS can achieve the state of the art results in LLL while requiring fewer parameters than other methods.

U2 - 10.1109/TPAMI.2022.3220928

DO - 10.1109/TPAMI.2022.3220928

M3 - Article

JO - IEEE Transactions on Pattern Analysis and Machine Intelligence

JF - IEEE Transactions on Pattern Analysis and Machine Intelligence

SN - 0162-8828

ER -