A transcription-based approach to determine the difficulty of a speech recognition task

A new parameter for estimating the difficulty of a continuous speech recogn ition task, call ed speech decoding difficulty, is presented in this work. It is obtained from the language model defined for the recognition task and the phonetic similarity between the transcriptions of the words that make up the vocabulary used. Two variants of the proposed task difficulty measur e are introduced: ideal speech decoding difficulty (ISDD), which is uninflu enced by practical considerations on the recognition system implemented, an d a second, more realistic variant, called practical speech decoding diffic ulty (PSDD) to study the performance of a specific recognition system when confronting a given task.