Unlock instant, AI-driven research and patent intelligence for your innovation.

Subtitle extraction method and device based on artificial intelligence, equipment and storage medium

An extraction method and extraction device technology, applied in the field of artificial intelligence, can solve the problems of affecting the accuracy of text recognition, not well mining the context relationship of subtitle sentences, and text recognition errors.

Pending Publication Date: 2022-04-15
PING AN TECH (SHENZHEN) CO LTD
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Since CTC assumes the conditional independence between the current output and the historical output, it does not dig out the contextual relationship in the subtitle sentence very well, so it may cause text recognition errors and affect the accuracy of text recognition

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
  • Subtitle extraction method and device based on artificial intelligence, equipment and storage medium
  • Subtitle extraction method and device based on artificial intelligence, equipment and storage medium
  • Subtitle extraction method and device based on artificial intelligence, equipment and storage medium

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0035] In order to make the object, technical solution and advantages of the present invention clearer, the present invention will be further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described here are only used to explain the present invention, not to limit the present invention.

[0036] It should be noted that although the functional modules are divided in the schematic diagram of the device, and the logical sequence is shown in the flowchart, in some cases, it can be executed in a different order than the module division in the device or the flowchart in the flowchart. steps shown or described. The terms "first", "second" and the like in the specification, claims or the above drawings are used to distinguish similar objects, and not necessarily used to describe a specific order or sequence.

[0037]In related technologies, the process of extracting video subtitles mainly ...

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 provides a subtitle extraction method and device based on artificial intelligence, equipment and a storage medium. The subtitle extraction method comprises the steps of obtaining a target textbox image; the target textbox image is input to a trained recognition model to recognize a target text in the target textbox image, the recognition model comprises a text image information extraction network, a language model information extraction network and a joint feedforward neural network, the text image information extraction network is obtained by pre-training a sample textbox image, and the language model information extraction network is obtained by pre-training a language model; the language model information extraction network is obtained by pre-training a sample video text, and the joint feedforward neural network is used for combining weight parameters of the two extraction networks after pre-training. According to the method, the text image information extraction result and the language model information extraction result can be combined through the joint feedforward neural network, the problem caused by CTC condition independence hypothesis is solved, image texture features and language features can be utilized during prediction, replacement errors are reduced, and the character recognition accuracy is improved.

Description

technical field [0001] Embodiments of the present invention relate to but are not limited to the technical field of artificial intelligence, and in particular, relate to an artificial intelligence-based subtitle extraction method, a subtitle extraction device, computer equipment, and a computer-readable storage medium. Background technique [0002] For the extraction process of video subtitles, it mainly includes text box position extraction and text recognition in the text box. Among them, the extraction of the position of the text box can be realized through the DB algorithm; in addition, for the text recognition in the text box, the current text detection and recognition methods usually adopt the more common CRNN and CTC methods. Since CTC assumes the conditional independence between the current output and the historical output, it does not dig out the contextual relationship in the subtitle sentence very well, so it may cause text recognition errors and affect the accura...

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): G06V30/414G06V20/40G06N3/04G06N3/08G06V30/10G06V10/82
CPCG06N3/08G06N3/044
Inventor 庞烨高欣建韩茂琨刘玉宇肖京
Owner PING AN TECH (SHENZHEN) CO LTD