Video label index database creation method and device and video label generation method and device

A video tag and index library technology is applied in the fields of video tag index library creation, video tag generation method and device, and can solve the problems of time-consuming, labor-intensive and poor scalability.

Active Publication Date: 2019-11-26
TENCENT TECH (SHENZHEN) CO LTD
View PDF7 Cites 7 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

It can be seen that the training process of this classification mode

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Video label index database creation method and device and video label generation method and device
  • Video label index database creation method and device and video label generation method and device
  • Video label index database creation method and device and video label generation method and device

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0081] The traditional video label generation method needs to use a large amount of training data to train the classification model. For a video that needs to generate a label, it is necessary to extract the features of the video, and then input the features into the trained classification model, which is analyzed by the classification model. The features of the video finally determine the label that matches the feature of the video. However, this method requires a large amount of training data to train the classification model. Moreover, once the classification model is trained, the type of labels it can classify will remain unchanged, such as When a new label is added, a large amount of training data corresponding to the new label is required to retrain the classification model. It can be seen that this kind of classification model requires a large amount of training data for training. The training process is time-consuming and laborious, and the scalability of the model is ve...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

PUM

No PUM Login to view more

Abstract

The invention discloses a video label index database creation method and device and a video label generation method and device, and the video label index database creation method comprises the steps:marking a certain number of videos as marked videos for each preset label; then, extracting a plurality of video frames meeting a preset condition from each labeled video, and extracting video frame features from each video frame; and for each preset label, obtaining a target video frame feature corresponding to the preset label based on the video frame feature of each video frame corresponding tothe preset label; and creating a video label index library according to each preset label and the corresponding target video frame feature. When the video label index database is created, each presetlabel only needs a small amount of labeled videos; moreover, when a new label is added, retraining is not needed, and only a small amount of labeled video extraction features corresponding to the newlabel need to be updated to the video label index database, so that the extension is easier.

Description

technical field [0001] The present application relates to the technical field of video processing, in particular to a method for creating a video tag index library, a method and a device for generating video tags. Background technique [0002] With the rapid development of network technology and the promotion and application of multimedia, various videos are continuously produced, and media data such as video have become the main body of big data. How to automatically generate relevant tags for a large number of videos will be of great significance to meet users' information acquisition needs. major. [0003] The traditional video label generation scheme is usually: divide the short video into multiple video clips, and obtain the key frames of each video clip; then, use the pre-trained classification model to classify each key frame to obtain the key frame label, according to the key The label corresponding to the frame gets the label of the video. Among them, the classifi...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

Application Information

Patent Timeline
no application Login to view more
IPC IPC(8): G06F16/71G06F16/75G06F16/78
CPCG06F16/71G06F16/75G06F16/7867
Inventor 李争郑茂
Owner TENCENT TECH (SHENZHEN) CO LTD
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products