Login for students

Login for employees

Publication detail

Refined Max-Pooling and Unpooling Layers for Deep Convolutional Neural Networks
Authors: Škrabánek Pavel
Year: 2016
Type of publication: článek ve sborníku
Name of source: Mendel 2016 : 22nd International Conference on Soft Computing
Publisher name: Vysoké učení technické v Brně
Place: Brno
Page from-to: 131-142
Titles:
Language Name Abstract Keywords
cze Vylepšené max-poolingové a unpoolingové vrstvy pro hluboké konvoluční sítě Příspěvek prezentuje nové pojetí poolingových a unpoolingových vrstev, které jsou vhodné pro hluboké konvoluční neuronové sítě. Za tímto účelem byla navržena nová aproximace inverzní funkce k funkci max-pooling. Rovněž bylo navrženo nové pojetí přepínačů, které jsou využívány k identifikaci polohy maxima. Koncepce byla testována na jednoduchém konvolučním auto-enkodéru.
eng Refined Max-Pooling and Unpooling Layers for Deep Convolutional Neural Networks The main goal of this paper is the introduction of new pooling and unpooling layers suited for deep convolutional neural networks. To this end, a new approximation of max-pooling inversion has been designed. The idea behind this approximation is also introduced in this paper. It is demonstrated on pools of size 2 x 2, with a stride of 2. The widely used technique of switches is combined with interpolation to form the new approximation. For that purpose, an unconventional expression of the switches has been used. Such an expression, allows the right placement of maxima in a reconstruction of original data, as well as interpolation of all unknown values in the reconstruction using the known maxima. The introduced inversion has been implemented into the aforementioned refined pooling and unpooling layers. Since they are suited for deep convolutional networks, behavior of the layers in the feed-forward and backpropagation passes had to be solved. In this context, the introduced conception of the switches has been further developed. Specifically, feed-forward and backpropagation switches are considered in the refined layers. One version of feed-forward and three versions of backpropagation switches have been introduced within this paper. The refined pooling and unpooling layers have been tested on a simple convolutional auto-encoder in order to verify functionality of the conception. convolutional neural network, unpooling, inversion of max-pooling, switches, deep learning, convolutional auto-encoder

Department

Map

en.mapy.cz

Services

Popular Links

© 2023 University of Pardubice, Studentská 95, 532 10 Pardubice 2
Phone: +420 466 036 111, 466 036 112, 466 036 113
Webmaster, RSS
Data box ID: f5vj9hu
Accessibility Statement
CC BY-NC-ND 4.0Creative CommonsAttributionNonCommercialNoDerivatives