Classification margin for improved class-based speech recognition performance

Denis Jouvet 1 Nicolas Vinuesa 1
1 PAROLE - Analysis, perception and recognition of speech
Inria Nancy - Grand Est, LORIA - NLPKD - Department of Natural Language Processing & Knowledge Discovery
Abstract : This paper investigates class-based speech recognition, and more precisely the impact of the selection of the training samples for each class on the final speech recognition performance. Increasing the number of recognition classes should lead to more specific models, and thus to better recognition performance, providing the trained model parameters are reliable. However, when the number of classes increases, the amount of training data for each class gets smaller, and may lead to unreliable parameters. The experiments described in the paper show that taking into account a classification margin tolerance helps associating more training data to each class, and improves the overall speech recognition performance.
Complete list of metadatas

https://hal.inria.fr/hal-00753345
Contributor : Denis Jouvet <>
Submitted on : Monday, November 19, 2012 - 10:04:31 AM
Last modification on : Tuesday, December 18, 2018 - 4:38:02 PM

Identifiers

Collections

Citation

Denis Jouvet, Nicolas Vinuesa. Classification margin for improved class-based speech recognition performance. ICASSP - 2012 - IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), Mar 2012, Kyoto, Japan. pp.4285 - 4288, ⟨10.1109/ICASSP.2012.6288866⟩. ⟨hal-00753345⟩

Share

Metrics

Record views

262