Online multimodal dictionary learning through Tucker decomposition

Abraham Traoré (1, 2) , Maxime Berar (2) , Alain Rakotomamonjy (1, 2)

1 LITIS - Laboratoire d'Informatique, de Traitement de l'Information et des Systèmes
2 UNIROUEN - Université de Rouen Normandie

We propose a new online approach for multimodal dictionary learning. The method developed in this work addresses the great challenges posed by the computational resource constraints in dynamic environment when dealing with large scale tensor sequences. Given a sequence of tensors, i.e. a set composed of equal-size tensors, the approach proposed in this paper allows to infer a basis of latent factors that generate these tensors by sequentially processing a small number of data samples instead of using the whole sequence at once. Our technique is based on block coordinate descent, gradient descent and recursive computations of the gradient. A theoretical result is provided and numerical experiments on both real and synthetic data sets are performed.

Mots clés: tensor; block coordinate descent; gradient descent; recursive computations; online dictionary learning

Domaines: Intelligence artificielle [cs.AI]

Abraham Traoré, Maxime Berar, Alain Rakotomamonjy. Online multimodal dictionary learning through Tucker decomposition. 2019. ⟨hal-01850923v3⟩