| 摘要 |
A picture composition method and apparatus for a speech recognition syntax tree, a terminal device and a storage medium. The method comprises: determining a groove corresponding to an application scene, and endowing each groove with at least one corresponding word (11); according to a groove sequence in a predetermined syntax path and the word corresponding to the groove, forming a syntax statement, analysing the speeches of the words in the syntax statement, and determining a corresponding acoustic model (12); and building a syntax tree according to the groove in the predetermined syntax path of the application scene, wherein the groove corresponds to a node in the syntax tree, and a groove index for the groove is stored in the node, the groove index being used for indexing an acoustic model storage position of the corresponding word (13). By means of the method, a picture composition process can be completed only by determining a groove corresponding to an application scene, endowing each groove with at least one corresponding word, building a syntax tree corresponding to the groove, and storing a groove index of the groove in a node, so that memory space and time consumed by picture composition can be greatly saved. |
