Conference paper

BASKAR Murali K., KARAFIÁT Martin, BURGET Lukáš, VESELÝ Karel, GRÉZL František and ČERNOCKÝ Jan. Residual Memory Networks: Feed-forward approach to learn long-term temporal dependencies. In: Proceedings of ICASSP 2017. New Orleans: IEEE Signal Processing Society, 2017, pp. 4810-4814. ISBN 978-1-5090-4117-6.
Publication language:english
Original title:Residual Memory Networks: Feed-forward approach to learn long-term temporal dependencies
Title (cs):Residuální paměťové sítě: nerekurentní přístup k učení dlouhých časových závislostí
Pages:4810-4814
Proceedings:Proceedings of ICASSP 2017
Conference:42nd IEEE International Conference on Acoustics, Speech and Signal Processing
Place:New Orleans, US
Year:2017
ISBN:978-1-5090-4117-6
Publisher:IEEE Signal Processing Society
URL:http://www.fit.vutbr.cz/research/groups/speech/publi/2017/baskar_icassp2017_0004810.pdf [PDF]
Files: 
+Type Name Title Size Last modified
iconbaskar_icassp2017_0004810.pdf245 KB2017-06-09 14:37:22
^ Select all
With selected:
Keywords
Automatic speech recognition, LSTM, RNN, Residual memory networks.
Annotation
This paper is about residual memory networks (RMN) structure, a variant of feed-forward network to capture temporal context.
Abstract
Training deep recurrent neural network (RNN) architectures is complicated due to the increased network complexity. This disrupts the learning of higher order abstracts using deep RNN. In case of feed-forward networks training deep structures is simple and faster while learning long-term temporal information is not possible. In this paper we propose a residual memory neural network (RMN) architecture to model short-time dependencies using deep feed-forward layers having residual and time delayed connections. The residual connection paves way to construct deeper networks by enabling unhindered flow of gradients and the time delay units capture temporal information with shared weights. The number of layers in RMN signifies both the hierarchical processing depth and temporal depth. The computational complexity in training RMN is significantly less when compared to deep recurrent networks. RMN is further extended as bi-directional RMN (BRMN) to capture both past and future information. Experimental analysis is done on AMI corpus to substantiate the capability of RMN in learning long-term information and hierarchical information. Recognition performance of RMN trained with 300 hours of Switchboard corpus is compared with various state-of-the-art LVCSR systems. The results indicate that RMN and BRMN gains 6 % and 3.8 % relative improvement over LSTM and BLSTM networks.
BibTeX:
@INPROCEEDINGS{
   author = {K. Murali Baskar and Martin Karafi{\'{a}}t and
	Luk{\'{a}}{\v{s}} Burget and Karel Vesel{\'{y}} and
	Franti{\v{s}}ek Gr{\'{e}}zl and Jan {\v{C}}ernock{\'{y}}},
   title = {Residual Memory Networks: Feed-forward approach to learn
	long-term temporal dependencies},
   pages = {4810--4814},
   booktitle = {Proceedings of ICASSP 2017},
   year = {2017},
   location = {New Orleans, US},
   publisher = {IEEE Signal Processing Society},
   ISBN = {978-1-5090-4117-6},
   language = {english},
   url = {http://www.fit.vutbr.cz/research/view_pub.php?id=11467}
}

Your IPv4 address: 54.162.60.75
Switch to IPv6 connection

DNSSEC [dnssec]