SOLUTIONS FOR ROBUST R~~O~~~TI~N OVER THE GSM CE
Lamia Karray, Aibdellatif Ben Jelloun, ChaJc Mokbel · 1998
This paper deals with automatic speech recognition robustness for noisy wireless communications. We propose several solutions to improve speech recognition aver the cellular network. Two architectures are derived far the recognizer. They are based on Hidden Markov Models (HMMs) adapted to adverse noise conditions. Then twio more specific solutions aiming to alleviate GSM cellular network defects (holes and impulsive noise) are developed. Holes are detected and rejected. Impulsive noises are modeled using mixture density HMMs and a maximum likelihood criterion. These solutions allow a noticeable recognition error reduction. The last one seems to be promising.