Title: Towards network simplification for low-cost devices by removing synapses
Authors: Bulín, Martin
Šmídl, Luboš
Švec, Jan
Citation: BULÍN, Martin; ŠMÍDL, Luboš; ŠVEC, Jan. Towards network simplification for low-cost devices by removing synapses. In: International conference on speech and computer: SPECOM 2018. Cham: Springer, 2018, p. 58-67. (Lectures notes in computer science; 11096). ISBN 978-3-319-99579-3.
Issue Date: 2018
Publisher: Springer
Document type: konferenční příspěvek
URI: http://hdl.handle.net/11025/34329
ISBN: 978-3-319-99579-3
Keywords: prořezávání synpsí;zjednodušení sítě;struktura minimální sítě;nízkonákladová zařízení;rozpoznávání řeči
Keywords in different language: pruning synapses;network simplification;minimal network structure;low-cost devices;speech recognition
Abstract in different language: The deployment of robust neural network based models on low-cost devices touches the problem with hardware constraints like limited memory footprint and computing power. This work presents a general method for a rapid reduction of parameters (80–90%) in a trained (DNN or LSTM) network by removing its redundant synapses, while the classification accuracy is not significantly hurt. The massive reduction of parameters leads to a notable decrease of the model’s size and the actual prediction time of on-board classifiers. We show the pruning results on a simple speech recognition task, however, the method is applicable to any classification data.
Rights: © Springer
Appears in Collections:Konferenční příspěvky / Conference Papers (KKY)

Files in This Item:
File Description SizeFormat 
Bulin.pdfPlný text1,6 MBAdobe PDFView/Open

Please use this identifier to cite or link to this item: http://hdl.handle.net/11025/34329

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.