Object detection model training method and device

Abstract

The invention relates to an object detection model training method and device, computer equipment and a storage medium. The method comprises the following steps: acquiring a sample image; through a to-be-trained object detection model, obtaining candidate areas where target objects are located in the sample images and initial confidence coefficients of the target objects in the candidate areas corresponding to the object categories; obtaining an attention response graph of the to-be-trained object detection model for the sample image according to the initial confidence of the target object corresponding to each object category in each candidate area and the gradient information between the image data of the sample image; obtaining an attention response value of each candidate region from the attention response graph, and adjusting network parameters of the to-be-trained object detection model according to the attention response value of each candidate region, the initial confidence coefficient of the target object corresponding to each object category in each candidate region and the object category label until a convergence condition is satisfied to obtain a target object detection model. By adopting the method, the detection performance of the object detection model can be improved.

Description

technical field [0001] The present application relates to the technical field of image processing, and in particular to an object detection model training method, device, computer equipment and storage medium, and an object detection method, device, computer equipment and storage medium. Background technique [0002] With the development of computer technology, it is one of the core issues of computer vision technology to detect and identify all the target objects of interest from the image through the object detection model, and to determine the area position of the target object in the image and the object category and other object information. . However, the target object often only occupies a small area in the image, while the background image other than the target object occupies most of the image, and the background image of a large area will cause the object detection model to detect the target from the image. Objects have an impact, resulting in a decrease in the ac...

AI Technical Summary

Problems solved by technology

However, the target object often only occupies a small area in the image, while the background image other than the target object occupies most of the image, and the background image of a large area will cause the object detection model to detect the target from the image. Objects have an impact, resulting in a decrease in the accuracy of detecting and identifying target objects from images

Images