Conference paper

HANNEMANN Mirko, POVEY Daniel and ZWEIG Geoffrey. Combining Forward and Backward Search in Decoding. In: Proceedings of ICASSP 2013. Vancouver: IEEE Signal Processing Society, 2013, pp. 6739-6743. ISBN 978-1-4799-0355-9.
Publication language:english
Original title:Combining Forward and Backward Search in Decoding
Title (cs):Kombinace dopředného a zpětného prohledávání v dekódování
Proceedings:Proceedings of ICASSP 2013
Conference:38th International Conference on Acoustics, Speech, and Signal Processing
Place:Vancouver, CA
Publisher:IEEE Signal Processing Society
speech decoding, beam width, search errors
This article describes a combination of forward and backward search in speech decoding based on WFST decoders.
We introduce a speed-up for weighted finite state transducer (WFST) based decoders, which is based on the idea that one decoding pass using a wider beam can be replaced by two decoding passes with smaller beams, decoding forward and backward in time. We apply this in a decoder that works with a variable beam width, which is widened in areas where the two decoding passes disagree. Experimental results are shown on the Wall Street Journal corpus (WSJ) using the Kaldi toolkit, and show a substantial speedup (a factor or 2 or 3) at the "more accurate" operating points. As part of this work we also introduce a new fast algorithm for weight pushing in WFSTs, and summarize an algorithm for the time reversal of backoff language models.
   author = {Mirko Hannemann and Daniel Povey and Geoffrey Zweig},
   title = {Combining Forward and Backward Search in Decoding},
   pages = {6739--6743},
   booktitle = {Proceedings of ICASSP 2013},
   year = {2013},
   location = {Vancouver, CA},
   publisher = {IEEE Signal Processing Society},
   ISBN = {978-1-4799-0355-9},
   language = {english},
   url = {}

Your IPv4 address:
Switch to IPv6 connection

DNSSEC [dnssec]