Automatic picture labeling method and device

Abstract

The invention discloses an automatic picture labeling method and device, relates to the technical field of picture processing, and aims to realize simultaneous labeling of multiple labels of a pictureand improve the accuracy of the labels while ensuring the labeling efficiency. The method comprises the following steps: training a multi-modal feature extraction model based on gallery data; constructing a visual semantic similarity nearest neighbor index of each picture and the label group according to the corresponding relationship between each picture and the label group and the classification; extracting features of a to-be-detected picture through a feature extraction model to obtain a feature vector, and matching a similar picture from the picture library data based on the feature vector and the visual semantic similarity nearest neighbor index; screening out an initial label of the to-be-detected picture according to the frequency and the weight of the keyword in the label group corresponding to the similar picture; and performing label filtering and weight sorting on the initial labels by adopting a pre-trained word vector model to obtain a final label group of the to-be-detected picture. The device applies the method provided by the scheme.

Description

technical field [0001] The invention relates to the technical field of picture processing, in particular to a method and device for automatic picture labeling. Background technique [0002] In recent years, with the extreme popularity of camera terminals, the vigorous development of self-media and the great abundance of Internet resources, how to effectively review, retrieve, extract and organize copyrighted media content has become a major challenge for the gallery industry. Keyword tags in image metadata play a key role in the review, retrieval and organization of massive images. However, the process of manually producing tags is cumbersome and complicated. Therefore, automatic tag generation has always been a key research field in computer vision and artificial intelligence. Image The automatic generation of tags has a wide range of application scenarios, which can effectively improve the labor efficiency of image editing and review, and optimize the structure of the prod...

AI Technical Summary

Problems solved by technology

Keyword tags in image metadata play a key role in the review, retrieval and organization of massive images. However, the process of manually producing tags is cumbersome and complicated. Therefore, automatic tag generation has always been a key research field in computer vision and artificial intelligence. Image The automatic generation of tags has a wide range of application scenarios, which can effectively improve the efficiency of image editing and review, and can optimize the structure of the original image production data from the media

[0003] Most of the existing image automatic labeling technologies are trained based on a single label model. To achieve multi-label labeling of pictures, it is necessary to train multiple types of label models to label them separately. Obviously, the existing automatic image labeling technology has low labeling efficiency. and poor accuracy

Images