Catastrophic forgetting (CF) happens whenever a neural network overwrites past knowledge while being trained on new tasks. Common techniques to handle CF include regularization of the weights (using, e.g., their importance on past tasks), and rehearsal strategies, where the network is constantly re-trained on past data. Generative models have also been applied for the latter, in order to have endless sources of data. In this paper, we propose a novel method that combines the strengths of regularization and generative-based rehearsal approaches. Our generative model consists of a normal-izing flow (NF), a probabilistic and invertible neural network, trained on the internal embeddings of the network. By keeping a single NF throughout the training process, we show that our memory overhead remains constant. In addition, exploiting the invertibility of the NF, we propose a simple approach to regularize the network's embeddings with respect to past tasks. We show that our method performs favorably with respect to state-of-the-art approaches in the literature, with bounded computational power and memory overheads.(c) 2023 Elsevier Ltd. All rights reserved.

Continual learning with invertible generative models / Pomponi, Jary; Scardapane, Simone; Uncini, Aurelio. - In: NEURAL NETWORKS. - ISSN 0893-6080. - 164:(2023), pp. 606-616. [10.1016/j.neunet.2023.05.020]

Continual learning with invertible generative models

Jary Pomponi;Simone Scardapane;Aurelio Uncini
2023

Abstract

Catastrophic forgetting (CF) happens whenever a neural network overwrites past knowledge while being trained on new tasks. Common techniques to handle CF include regularization of the weights (using, e.g., their importance on past tasks), and rehearsal strategies, where the network is constantly re-trained on past data. Generative models have also been applied for the latter, in order to have endless sources of data. In this paper, we propose a novel method that combines the strengths of regularization and generative-based rehearsal approaches. Our generative model consists of a normal-izing flow (NF), a probabilistic and invertible neural network, trained on the internal embeddings of the network. By keeping a single NF throughout the training process, we show that our memory overhead remains constant. In addition, exploiting the invertibility of the NF, we propose a simple approach to regularize the network's embeddings with respect to past tasks. We show that our method performs favorably with respect to state-of-the-art approaches in the literature, with bounded computational power and memory overheads.(c) 2023 Elsevier Ltd. All rights reserved.
2023
catastrophic forgetting; continual learning; machine learning; normalizing flow
01 Pubblicazione su rivista::01a Articolo in rivista
Continual learning with invertible generative models / Pomponi, Jary; Scardapane, Simone; Uncini, Aurelio. - In: NEURAL NETWORKS. - ISSN 0893-6080. - 164:(2023), pp. 606-616. [10.1016/j.neunet.2023.05.020]
File allegati a questo prodotto
File Dimensione Formato  
Pomponi_Continual learning_2023.pdf

accesso aperto

Tipologia: Versione editoriale (versione pubblicata con il layout dell'editore)
Licenza: Creative commons
Dimensione 774 kB
Formato Adobe PDF
774 kB Adobe PDF

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11573/1686918
Citazioni
  • ???jsp.display-item.citation.pmc??? 0
  • Scopus 1
  • ???jsp.display-item.citation.isi??? 0
social impact