Target object detection method, device, computer equipment and storage medium

Abstract

The present application relates to a target object detection method, device, computer equipment, and storage medium, including: inputting a training image including a label frame into a target object detection model to be trained; Perform feature processing on the feature map to obtain the initial prediction labeling frame corresponding to each preset labeling frame on the initial training feature map, and determine the update labeling frame from each initial prediction labeling frame based on the position difference between the preset labeling frame and the label labeling frame; Transform and fuse the training feature map to obtain the target training feature map; perform feature processing on the target training feature map based on the updated label box, and obtain the target prediction label box corresponding to the updated label box; based on the initial prediction label box, the target prediction label box and the label The difference in the position of the label box generates a regression loss, and the target object detection model is trained based on the regression loss, so as to train a more accurate target object detection model.

Description

technical field [0001] The present application relates to the field of computer technology, in particular to a target object detection method, device, computer equipment and storage medium. Background technique [0002] With the development of computer technology, face detection technology has emerged. Face detection plays an important role in many fields such as video surveillance, social entertainment, access control check-in, facial expression analysis, and image search. [0003] The traditional face detection algorithm uses manually designed features to extract features from the image, and then uses a sliding window to perform template matching on the image to be detected. If the features of the image area and the face template highly match, the area is considered to be the face area. The manual feature part of the traditional face detection algorithm has limited generalization performance, which leads to low detection accuracy. Contents of the invention [0004] Base...

AI Technical Summary

Problems solved by technology

This traditional face detection algorithm has limited generalization performance of manual features, resulting in low detection accuracy

Images