Command word recognition method and system using mutual prior constraint model

The invention discloses a command word recognition method using a mutual prior constraint model. The command word recognition method is based on an end-to-end voice command word recognition structure,wherein the end-to-end voice command word recognition structure comprises a phoneme module used for extracting phoneme features from audio, a word module used for extracting word features from the phoneme features, and a semantic module used for extracting semantic features from the word features. The command word recognition method is characterized in that the semantic module recognizes commandwords according to a sequential order, changes an independent recognition process into a sequential recognition process with mutual prior constraints, i.e., recognizes a content component firstly, andthen influence recognition results of other content components through using the content component, so that the purpose of improving the accuracy rate is achieved. The invention further provides a corresponding system. Compare