An End-to-End Structured Method for News Programs

A structured and news technology, applied in speech analysis, instrument, character and pattern recognition, etc., can solve the problem of unsatisfied real-time performance of the secondary editing of news programs, and improve the timeliness of use, reduce errors, and improve accuracy. rate effect

CN110012349BActive Publication Date: 2019-09-20CHENGDU SOBEY DIGITAL TECH CO LTD
11 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Publication Date
2019-09-20

Smart Images

  • Figure 1
    Figure 1
  • Figure 2
    Figure 2
Patent Text Reader

Abstract

The invention discloses an end-to-end news program structuring method and its structuring framework system, and relates to the technical field of news program processing. The method of the invention includes preprocessing the input news program, and obtaining the audio resource and video of the news program resources; use ASR speech recognition technology, OCR text recognition technology and Shot Detection technology to extract basic information in audio resources and video resources; based on the extracted basic information, extract semantic theme information of each modality, and adopt cross-modal theme fusion Extraction algorithm, fusion and clustering of semantic theme information of each modality, and output Scene theme; at the same time, cross-modal scene detection, output Scene level; use CRF scene labeling algorithm to perform scene aggregation and segmentation on the obtained Scene level and Scene theme , and output Story levels and Story topics with the same semantics. The present invention focuses on Story layers and Scene layers with clear semantic meanings, which facilitates the secondary utilization of news programs and improves the timeliness of news programs.
Need to check novelty before this filing date? Find Prior Art

Description

technical field

[0001] The present invention relates to the technical field of news program processing, and more specifically relates to an end-to-end news program structuring method and its structuring framework system. Background technique

[0002] With the development of the times and the advancement of technology, video indexing and retrieval is an important issue and has great significance. And TV news is a large part of the video, and it is also a video that will be used repeatedly. For example, for on-demand broadcasting of TV news, it is necessary to segment the TV news stream, and then mark each segment of TV news stream with metadata, so as to quickly index and access; TV news programs are used as a material again, and used as The editorial materials of other news programs often reuse valuable fragments of news, and it is also necessary to decompose the TV news stream according to the structure of TV news and mark the valuable fragments.

[0003] News video is an...

Examples

Embodiment 1

[0030] Such as figure 1 and figure 2 As shown, this embodiment provides an end-to-end news program structuring method, including the following steps:

[0031] S1: Preprocessing the input news program, respectively obtaining audio resources and video resources of the news program;

[0032] S2: Use ASR speech recognition technology, OCR text recognition technology and lens detection technology to extract basic information in audio resources and video resources;

[0033] Performing basic information analysis on audio resources in said S2 includes:

[0034] Based on the audio judgment analysis technology of MFCCs audio characteristics, the sound information of audio resources can be identified, and the pause interval of speech broadcast can be judged through audio feature analysis. The audio pause interval time points will be used for subsequent scene detection;

[0035] The voice content of audio resources is converted into text content through ASR speech recognition technolo...