A predictive speech encoder using coding scheme selection patterns to reduce sensitivity to frame errors

摘要

"A PREDICTIVE SPEECH TALK ENCODER USING CODING SCHEME SELECTION STANDARDS TO REDUCE SENSITIVITY TO FRAMEWORK ERRORS. A method and equipment for using coding scheme selection patterns in a predictive speech encoder to reduce sensitivity to speech conditions. frame errors includes a speech encoder configured to select from several predictive encoding modes.After a predefined number of speech frames have been predictively encoded, the speech encoder encodes a frame with either a non-predictive encoding mode or an encoding mode slightly predictive. The preset number of frames can be determined in advance from a listener's subjective point of view. The preset number of frames can be varied periodically. An average encoding bit rate can be maintained for the speech encoder ensuring an average encoding bit rate is maintained for each the standard, or group, successively of predictively encoded speech frames, including at least one non-predictively encoded or slightly predictive encoded speech frame.