Word Error Rate
Word Error Rate is a quantitative evaluation metric used in speech recognition that measures the proportion of incorrectly recognized words relative to a reference transcript, calculated as the sum of substitutions, deletions, and insertions divided by the total number of words in the reference.
1.7K
Publications
91.4K
Citations
3.6K
Authors
743
Institutions
Publications per year
2017–2026
488
Parents
1.7K
Recurrent neural network based language model
Tomáš Mikolov, Martin Karafiát, Lukáš Burget et al. · 2010 · 5.4K citations
Hybrid speech recognition with Deep Bidirectional LSTM
Alex Graves, Navdeep Jaitly, Abdelrahman Mohamed · 2013 · 1.8K citations
Improved backing-off for M-gram language modeling
Reinhard Kneser, Hermann Ney · 2002 · 1.5K citations
1–5 of 1.7K
3.6K
34
h-index
25
h-index
20
h-index
20
h-index
IBM Research - Thomas J. Watson Research Center, United States
21 publications · 1.3K citations
20
h-index
1–5 of 3.6K
743
Carnegie Mellon University
Pittsburgh, United States
38 h-index
224 publications
12.1K citations
Google (United States)
Mountain View, United States
35 h-index
356 publications
32.1K citations
RWTH Aachen University
Aachen, Germany
34 h-index
177 publications
13K citations

Johns Hopkins University
Baltimore, United States
32 h-index
136 publications
12.3K citations
IBM Research - Thomas J. Watson Research Center
Yorktown Heights, United States
32 h-index
218 publications
10.2K citations
Microsoft (United States)
Redmond, United States
30 h-index
201 publications
9.7K citations
IBM (United States)
Armonk, United States
30 h-index
167 publications
11.5K citations
University of Cambridge
Cambridge, United Kingdom
28 h-index
149 publications
6.4K citations
1–8 of 743
35 publications · 3.3K citations
27
h-index
Repository · 62 publications · 2.5K citations
24
h-index
TASLP · 33 publications · 2.3K citations
21
h-index
30 publications · 1.4K citations
21
h-index
Conference · 17 publications · 673 citations
14
h-index
1–5