Classification margin for improved class-based speech recognition performance - Archive ouverte HAL Access content directly
Conference Papers Year : 2012

Classification margin for improved class-based speech recognition performance

(1) , (1)
1

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.
Not file

Dates and versions

hal-00753345 , version 1 (19-11-2012)

Identifiers

Cite

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⟩
115 View
0 Download

Altmetric

Share

Gmail Facebook Twitter LinkedIn More