Large vocabulary continuous speech recognition using HTK

By Philip C. Woodland, J. Odell, V. Valtchev, Susannah Young. 1994

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Speech & Audio

Citations in the corpus

wav2vec: Unsupervised Pre-training for Speech Recognition (2019, neutral): “Wall Street Journal (WSJ; (Garofolo et al., 1993a; Woodland et al., 1994)) comprises about 81 hours of transcribed audio data.”
Convolutional Neural Networks-based continuous speech recognition using raw speech signal (2015, methodology): “The SI-284 set of the Wall Street Journal (WSJ) corpus [20] is formed by combining data from WSJ0 and WSJ1 databases, sampled at 16 kHz.”
CAPE: Encoding Relative Positions with Continuous Augmented Positional Embeddings (2021, methodology): “For WSJ data we consider the standard subsets si284 , nov93dev and nov92 for training, validation and test, respectively.”
Automatic speech recognition and speech variability: A review (2007, methodology): “Gender dependence is applied to whole word units, for example digits [101], or to context dependent phonetic units [224], as a result of an adequate splitting of the training data.”
End-to-end Phoneme Sequence Recognition using Convolutional Neural Networks (2013, methodology): “2 Wall Street Journal Corpus The SI-284 set of the corpus [26] is selected for the experiments.”
The Kaldi Speech Recognition Toolkit (2011, methodology): “The baseline results are reported in [24], which we refer to as “Bell” (for Bell Labs, the authors’ affiliation), and a HTK system described in [25].”
Speech Recognition (2018, unknown): “Wall Street Journal (WSJ; Garofolo et al. (1993a); Woodland et al. (1994)) comprises about 81 hours of transcribed audio data.”
Who Needs Words? Lexicon-Free Speech Recognition (2019, methodology): “Experiments We experiment with the Wall Street Journal (WSJ) dataset [28] (about 81 hours of transcribed audio data) and the Librispeech dataset [29] (1000 hours with clean and noisy speech).”