method and device for reducing the effects of transmission errors in a distributed speech recognition system and u0440u0430u0441u043fu0440u0435u0434u0435u043bu0435u043du043du043eu043c process

摘要

proposed method and device to reduce the impact of error transfer services for distributed speech recognition.the way is that the identified parameters and errors of speech recognition, and process the data intended for the to the decoder in speech recognitionin order to guarantee the removal of the final processing of all parameters for speech recognition, which have been mistakes.the parameters of speech recognition errors, which were excluded from the data transfer, by changing the parameters of the data selected,to u043eu0442u0431u0440u0430u043au043eu0432u044bu0432u0430u043bu0438u0441u044c decoder speech recognition as abnormal and / or data, such u0440u0435u0447u0438.u043cu0435u0436u0434u0443u043du0430u0440u043eu0434u043du0430u044f application was published, together with the report of the meeting u0435u0436u0434u0443u043du0430u0440u043eu0434u043du043eu043c search.