Human-computer interaction identity intelligent recognition method and system based on multi-modal perception

Abstract

The application provides a kind of man-machine interaction identity intelligent identification method and system based on multi-modal perception, it is related to man-machine interaction technical field, first capture the behavior signal hierarchy sequence of interactive subject in man-machine interaction, trigger multi-dimensional perception feature derivation transmission generation feature derivation transmission unit, the feature derivation transmission unit contains the derivation and transmission relationship data of physiological, behavior mode, response feedback and the like feature, then according to progressive order unit chain is linked to form feature derivation transmission time sequence chain, record feature dynamic evolution track, then call preset identity benchmark chain library, combine scene parameter and historical data dynamically update generation scene adaptation benchmark chain, then extract cross-dimension core matching feature, with dynamic weight distribution strategy bidirectional track alignment, finally based on alignment result generation contains identity attribution and the like information of identity recognition conclusion, the method improves identity recognition accuracy, adaptability and intelligent level.

Images