Importance order ranking for texture extraction. A more efficient pooling operator than max pooling? - Université d'Évry
Communication Dans Un Congrès Année : 2022

Importance order ranking for texture extraction. A more efficient pooling operator than max pooling?

Résumé

Much of convolutional neural network (CNN)'s success lies in translation invariance. The other part resides in the fact that thanks to a judicious choice of architecture, the network is able to make decisions taking into account the whole image. This work provides an alternative way to extend the pooling function, we named rank-order pooling, capable of extracting texture descriptors from images. The rank-order pooling layers are non parametric, independent of the geometric arrangement or sizes of the image regions, and can therefore better tolerate rotations. Rank-order pooling functions produce images capable of emphasizing low/high frequencies, contours, etc. We shows rank-order pooling leads to CNN models which can optimally exploit information from their receptive field. a
Fichier principal
Vignette du fichier
ICINCO_2022_21_CR.pdf (1.87 Mo) Télécharger le fichier
Origine Fichiers produits par l'(les) auteur(s)

Dates et versions

hal-04367045 , version 1 (29-12-2023)

Identifiants

Citer

Sofia Vargas Ibarra, Vincent Vigneron, Jean-Philippe Congé, Hichem Maaref. Importance order ranking for texture extraction. A more efficient pooling operator than max pooling?. 19th International Conference on Informatics in Control, Automation and Robotics (ICINCO 2022), Jul 2022, Lisbon, Portugal. pp.585--594, ⟨10.5220/0011142200003271⟩. ⟨hal-04367045⟩
32 Consultations
25 Téléchargements

Altmetric

Partager

More