Compressing the Input for CNNs with the ...
Type de document :
Communication dans un congrès avec actes
Titre :
Compressing the Input for CNNs with the First-Order Scattering Transform
Auteur(s) :
Oyallon, Edouard [Auteur]
Centre de vision numérique [CVN]
Sequential Learning [SEQUEL]
Organ Modeling through Extraction, Representation and Understanding of Medical Image Content [GALEN-POST]
Belilovsky, Eugene [Auteur]
Département d'Informatique et de Recherche Opérationnelle [Montreal] [DIRO]
Zagoruyko, Sergey [Auteur]
Models of visual object recognition and scene understanding [WILLOW]
Valko, Michal [Auteur]
Sequential Learning [SEQUEL]
Centre de vision numérique [CVN]
Sequential Learning [SEQUEL]
Organ Modeling through Extraction, Representation and Understanding of Medical Image Content [GALEN-POST]
Belilovsky, Eugene [Auteur]
Département d'Informatique et de Recherche Opérationnelle [Montreal] [DIRO]
Zagoruyko, Sergey [Auteur]
Models of visual object recognition and scene understanding [WILLOW]
Valko, Michal [Auteur]
Sequential Learning [SEQUEL]
Titre de la manifestation scientifique :
ECCV 2018 - European Conference on Computer Vision
Ville :
Munich
Pays :
Allemagne
Date de début de la manifestation scientifique :
2018-09-08
Mot(s)-clé(s) en anglais :
First-order scattering
SIFT
CNN
Image descriptors
SIFT
CNN
Image descriptors
Discipline(s) HAL :
Statistiques [stat]/Machine Learning [stat.ML]
Résumé en anglais : [en]
We study the first-order scattering transform as a candidate for reducing the signal processed by a convolutional neural network (CNN). We study this transformation and show theoretical and empirical evidence that in the ...
Lire la suite >We study the first-order scattering transform as a candidate for reducing the signal processed by a convolutional neural network (CNN). We study this transformation and show theoretical and empirical evidence that in the case of natural images and sufficiently small translation invariance, this transform preserves most of the signal information needed for classification while substantially reducing the spatial resolution and total signal size. We show that cascading a CNN with this representation performs on par with ImageNet classification models commonly used in downstream tasks such as the ResNet-50. We subsequently apply our trained hybrid ImageNet model as a base model on a detection system, which has typically larger image inputs. On Pascal VOC and COCO detection tasks we deliver substantial improvements in the inference speed and training memory consumption compared to models trained directly on the input image.Lire moins >
Lire la suite >We study the first-order scattering transform as a candidate for reducing the signal processed by a convolutional neural network (CNN). We study this transformation and show theoretical and empirical evidence that in the case of natural images and sufficiently small translation invariance, this transform preserves most of the signal information needed for classification while substantially reducing the spatial resolution and total signal size. We show that cascading a CNN with this representation performs on par with ImageNet classification models commonly used in downstream tasks such as the ResNet-50. We subsequently apply our trained hybrid ImageNet model as a base model on a detection system, which has typically larger image inputs. On Pascal VOC and COCO detection tasks we deliver substantial improvements in the inference speed and training memory consumption compared to models trained directly on the input image.Lire moins >
Langue :
Anglais
Comité de lecture :
Oui
Audience :
Internationale
Vulgarisation :
Non
Collections :
Source :
Fichiers
- https://hal.inria.fr/hal-01850921/document
- Accès libre
- Accéder au document
- https://hal.inria.fr/hal-01850921/document
- Accès libre
- Accéder au document
- https://hal.inria.fr/hal-01850921/document
- Accès libre
- Accéder au document
- document
- Accès libre
- Accéder au document
- ECCV2018.pdf
- Accès libre
- Accéder au document