Deep representation design from deep kernel networks - CNRS - Centre national de la recherche scientifique Accéder directement au contenu
Article Dans Une Revue Pattern Recognition Année : 2019

Deep representation design from deep kernel networks

Résumé

Deep kernel learning aims at designing nonlinear combinations of multiple standard elementary kernels by training deep networks. This scheme has proven to be effective, but intractable when handling large-scale datasets especially when the depth of the trained networks increases; indeed, the complexity of evaluating these networks scales quadratically w.r.t. the size of training data and linearly w.r.t. the depth of the trained networks. In this paper, we address the issue of efficient computation in Deep Kernel Networks (DKNs) by designing effective maps in the underlying Reproducing Kernel Hilbert Spaces (RKHS). Given a pretrained DKN, our method builds its associated Deep Map Network (DMN) whose inner product approximates the original network while being far more efficient. The design principle of our method is greedy and achieved layer-wise, by finding maps that approximate DKNs at different (input, intermediate and output) layers. This design also considers an extra fine-tuning step based on unsupervised learning, that further enhances the generalization ability of the trained DMNs. When plugged into SVMs, these DMNs turn out to be as accurate as the underlying DKNs while being at least an order of magnitude faster on large-scale datasets, as shown through extensive experiments on the challenging ImageCLEF, COREL5k benchmarks and the Banana dataset.
Fichier principal
Vignette du fichier
paper (1).pdf (1.4 Mo) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)

Dates et versions

hal-02325793 , version 1 (18-12-2020)

Identifiants

Citer

Mingyuan Jiu, Hichem Sahbi. Deep representation design from deep kernel networks. Pattern Recognition, 2019, 88, pp.447-457. ⟨10.1016/j.patcog.2018.12.005⟩. ⟨hal-02325793⟩
67 Consultations
91 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More