Human action intention recognition training method based on cooperative computation of multiple brain areas

Abstract

The invention belongs to the field of cognitive neuroscience, and specifically relates to a human action intention recognition training method based on cooperative computation of multiple brain areas.The human action intention recognition training method comprises the steps of 1, performing image collection on a human body action; 2, obtaining human body joint information, and recognizing the category of the human body action; 3, calculating a robot action strategy according to the category of the action executed by the human by adopting a mode of cooperative computation of multiple brain areas based on a brain-like computing model; 4, inputting a correctness judgment for the robot action strategy calculated in the step 3; 5, adjusting parameters of the brain-like computing model throughan STDP mechanism based on the correctness judgment inputted in the step 4; and 6, if the correctness judgment inputted in the step 4 shows that the robot action strategy is wrong, executing the step1 for repeated training until the correctness judgment inputted in the step 4 shows that the robot action strategy is correct. The human action intention recognition training method overcomes the defect of being not flexible enough because programming and the like need to be performed in advance in the traditional human-computer interaction technology, and improves the use experience.

Description

technical field [0001] The invention belongs to the field of cognitive neuroscience, and in particular relates to a human action intention recognition training method based on multi-brain area collaborative calculation. Background technique [0002] With the continuous development of artificial intelligence technology and robot manufacturing technology, robots are gradually integrated into human's daily life. In order to enable robots to better serve humans, adaptable, simple and flexible human-computer interaction technology is essential. Most of the human-computer interaction technologies currently applied to home service robots are programmed by manufacturers first, and then guide users to trigger them through voice commands or gesture commands. Although these methods are simple and effective, they lack a certain degree of flexibility, especially when the programming program is contrary to the user's habits, which will lead to a sharp decline in user experience satisfact...

AI Technical Summary

Problems solved by technology

[0004] In order to solve the above-mentioned problems in the prior art, that is, in order to solve the problem of insufficient flexibility of the human-computer interaction of the robot, the present invention proposes a human action intention recognition training method based on multi-brain region collaborative computing, including the following steps:

Images