Cross-domain grabbing recognition method and device, electronic equipment and storage medium

Abstract

The invention discloses a cross-domain capture recognition method and device, electronic equipment and a storage medium, and relates to the technical field of vision, and the cross-domain capture recognition method comprises the steps: generating an image cross-domain generative adversarial network; generating a pseudo target domain image with a target domain style from a source domain image by using the image cross-domain generative adversarial network; constructing an initial grabbing recognition model; pre-training the initial capture recognition model by using the pseudo target domain image; obtaining a target object cross-domain grabbing recognition model according to a preset domain classifier and the initial grabbing recognition model; and establishing a multi-target object capturing data set, and performing cross-domain capturing recognition of the target object in the complex scene according to the multi-target object capturing data set and the target object cross-domain capturing recognition model. According to the cross-domain capture recognition method, the difference of different-domain data can be reduced from the data level and the network structure level, and effective cross-domain migration of the capture recognition model is achieved.

Description

Technical field [0001] The present invention relates to the field of visual techniques, in particular, to a cross-domain grab recognition method, apparatus, electronic device, and storage medium. Background technique [0002] The capture operation serves as a basic function of the robot and plays a crucial role in promoting the intelligent road of robots. The technology has been widely received, especially in recent years, deep learning has achieved breakthrough results in the field of object testing, scene understanding, style migration, and launched a large number of robot grabbing recognition algorithms in deep learning methods. And achieved fruitful results. [0003] The robot grab recognition algorithm is mainly for simple environments, some methods, despite the measurement of multi-object gripping position in complex scenes, two-step or even multi-step identification methods are used to measure the grasp position measurement. The process changed, which reduces the recogniti...

AI Technical Summary

Problems solved by technology

[0003] The robot grasping recognition algorithm is mainly aimed at the grasping and recognition of objects in a simple environment. Although some methods realize the measurement of the grasping pose of multiple objects in complex scenes, they use a two-step or even multi-step recognition method to make the grasping pose measurement The process becomes cumbersome and reduces the recognition speed

On the basis of the traditional recognition algorithm, the deep learning-based robot grasping recognition algorithm has achieved significant advantages in the recognition effect, but it requires a large number of labeled training samples, and the cost of manpower and material resources to obtain these samples is very expensive. , and the cycle is long

Images