The MALACH corpus is packaged with benchmark train/test partitions and a 21.7% WER baseline using LSTM acoustic and language models.
This work would not be possible without the support of Sam Gustman from the Visual History Foundation
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2019 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Challenging the Boundaries of Speech Recognition: The MALACH Corpus
The MALACH corpus is packaged with benchmark train/test partitions and a 21.7% WER baseline using LSTM acoustic and language models.