AI video cloud interaction system based on simulated person logic knowledge base

Abstract

The invention discloses an AI video cloud interaction system based on a simulated person logic knowledge base. The system integrates an interruption interaction method, a voice shutdown method and a language correction method. The interruption interaction method realizes the controllability of an interruption instruction, interruption times, interaction interruption and awakening of man-machine interaction; according to the voice shutdown method, a power-off voice instruction is added in a system, and shutdown of the system is directly controlled through a voice instruction; and according to the language correction method, the function of pre-judging and correcting vocabularies with similar vocabularies and different characters is added in a system, and even if a large number of vocabularies exist in the interaction process, a corresponding mechanism of converting characters into pinyin can be rapidly achieved. According to the method, the interruption interaction method, the voice shutdown method and the language correction method are integrated, so that robot interaction is more human-like and intelligent, use of a user and commercialized popularization are greatly facilitated, the intelligent interaction effect is remarkably improved in the aspects of accuracy and judgment ability, and the user experience is further enhanced.

Description

Technical field [0001] The present invention belongs to the field of intelligent systems, and more particularly to a AI video cloud interaction system based on the experimental logic knowledge base. Background technique [0002] Existing video interactive systems, each time the user interacts, interrupts the current speech of the robot, not close enough to the true people's dialogue simulation, once the interaction begins, uninterrupted until the system is shut down. [0003] And the existing AI video interactive system is generally deployed in the Linux system, the shutdown method is to use manually entered the Linux command, and the operation interface is hidden under the playback interface, the operation is cumbersome, the use threshold is high, and it is not suitable for commercialization. [0004] And the existing video interaction system is influenced by the objective variables such as the region, habit, pronunciation method of the user, so that the standard degree of Manda...

AI Technical Summary

Problems solved by technology

[0002] The existing video interaction system will interrupt the robot's current speech every time the user interacts, which is not close enough to the real person's dialogue simulation, and once the interaction starts, it cannot be interrupted until the system is shut down

[0003] Moreover, the existing AI video interactive system is generally deployed in the linux system. The shutdown method is to manually input linux commands, and the operation interface is hidden under the playback interface. The operation is cumbersome and the threshold for use is high, which is not suitable for commercial promotion.

[0004] Moreover, when the existing video interactive system analyzes the user's language, due to the influence of objective variables such as the user's location, habits, and pronunciation methods, the standard of Mandarin varies, and the language-to-text recognition program cannot be 100% accurate. , combined with other situations such as the interference of network transmission on the audio stream, resulting in words with similar sounds but different words in the analyzed keywords, which greatly affects the correct judgment of the video interactive system program

Images