Publication Details
Generating Exact Lattices in The WFST Framework
Hannemann Mirko, Dipl.-Ing. (DCGM FIT BUT)
Boulianne Gilles (CRIM)
Burget Lukáš, doc. Ing., Ph.D. (DCGM FIT BUT)
Ghoshal Arnab (UEDIN)
Janda Miloš, Ing. (DCGM FIT BUT)
Karafiát Martin, Ing., Ph.D. (DCGM FIT BUT)
Kombrink Stefan, Dipl.-Inf -Ling (DCGM FIT BUT)
Motlíček Petr, doc. Ing., Ph.D. (DCGM FIT BUT)
Qian Yanmin (SJTU)
Riedhammer Korbinian (ICSI Berkeley)
Veselý Karel, Ing., Ph.D. (DCGM FIT BUT)
Vu Ngoc Thang (KIT)
Speech Recognition, Lattice Generation
This article describes a lattice generation method, which is most directly applicable when using WFST decoders.
We describe a lattice generation method that is exact, i.e. it satisfies all the natural properties we would want from a lattice of alternative transcriptions of an utterance. This method does not introduce substantial overhead above one-best decoding. Our method is most directly applicable when using WFST decoders where the WFST is "fully expanded", i.e. where the arcs correspond to HMM transitions. It outputs lattices that include state-level alignments as well as word labels. The general idea is to create a state-level lattice during decoding, and to do a special form of determinization that retains only the best-scoring path for each word sequence. This special determinization algorithm is a solution to the following problem: Given a WFST A, compute a WFST B that, for each input-symbolsequence of A, contains just the lowest-cost path through A. Our solution always works if the projection of A on its input labels is determinizable, e.g. if A is acyclic.
@INPROCEEDINGS{FITPUB9914,
author = "Daniel Povey and Mirko Hannemann and Gilles Boulianne and Luk\'{a}\v{s} Burget and Arnab Ghoshal and Milo\v{s} Janda and Martin Karafi\'{a}t and Stefan Kombrink and Petr Motl\'{i}\v{c}ek and Yanmin Qian and Korbinian Riedhammer and Karel Vesel\'{y} and Thang Ngoc Vu",
title = "Generating Exact Lattices in The WFST Framework",
pages = "4213--4216",
booktitle = "Proceedings of 2012 IEEE International Conference on Acoustics, Speech and Signal Processing",
year = 2012,
location = "Kyoto, JP",
publisher = "IEEE Signal Processing Society",
ISBN = "978-1-4673-0044-5",
doi = "10.1109/ICASSP.2012.6288848",
language = "english",
url = "https://www.fit.vut.cz/research/publication/9914"
}