An investigation of likelihood normalization for robust ASR

Abstract : Noise-robust automatic speech recognition (ASR) systems rely on feature and/or model compensation. Existing compensation techniques typically operate on the features or on the parameters of the acoustic models themselves. By contrast, a number of normalization techniques have been defined in the field of speaker verification that operate on the resulting log-likelihood scores. In this paper, we provide a theoretical motivation for likelihood normalization due to the so-called "hubness" phenomenon and we evaluate the benefit of several normalization techniques on ASR accuracy for the 2nd CHiME Challenge task. We show that symmetric normalization (S-norm) reduces the relative error rate by 43% alone and by 10% after feature and model compensation.
Complete list of metadatas

Cited literature [35 references]  Display  Hide  Download

https://hal.inria.fr/hal-01006142
Contributor : Emmanuel Vincent <>
Submitted on : Friday, June 13, 2014 - 9:17:46 PM
Last modification on : Saturday, March 30, 2019 - 1:26:31 AM
Long-term archiving on : Saturday, September 13, 2014 - 11:40:29 AM

File

vincent_IS14.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : hal-01006142, version 1

Citation

Emmanuel Vincent, Aggelos Gkiokas, Dominik Schnitzer, Arthur Flexer. An investigation of likelihood normalization for robust ASR. Interspeech, Sep 2014, Singapore, Singapore. ⟨hal-01006142⟩

Share

Metrics

Record views

448

Files downloads

570