Sound source localization processing method and device based on multi-modal information fusion

The invention relates to the technical field of intelligent electric appliances, and provides a sound source positioning processing method and device based on multi-modal information fusion, and the method comprises the steps: collecting image data of a position where a target sound source object is located through an image collection device in intelligent equipment, determining a correction weight vector corresponding to a microphone sensor array based on a relative horizontal distance relationship between the position of the target sound source object and the distribution position of the microphone sensor array in the intelligent device; obtaining a target weight vector of the microphone sensor array based on an initial weight vector and a corrected weight vector corresponding to the microphone sensor array in the intelligent equipment, and determining the relative position of the target sound source object and each microphone sensor based on the target weight value corresponding to each microphone sensor i