基于人工智能的智慧媒资典型应用与实践

Typical Application and Practice of Intelligent Media Asset Management System Based on Artificial Intelligence

电视台媒体资产管理系统是一直以馆藏型资料库存在,通常情况下媒体资产的应用效率极其低下。为了提高媒资素材的使用效率及历史资料创造新价值与效率,可以应用了人智工智能技术,打造了一套智慧型媒资,让馆藏型媒资库华丽转身,为其融媒体的发展提供重要的资源支撑。本文介绍了如何使用人工智能技术,在传统媒资的资源入库、资源编目、场景切分及智能检索等方面为媒体资产赋予新的活力,提升传统媒资的在业务生产过程中的作用。并且经过实践验证,在人脸识别、语音识别、OCR字幕识别等方面,已经达到目前智慧媒资的先进水平,最终也达到了应用要求,也为广电业务生产及智慧资源应用提供参考。

Media asset management system has always existed as a collection-type database in TV station.Generally,its application efficiency is extremely low.In order to improve use efficiency of media materials and create new value of historical materials,artificial intelligence technology can be applied to create a set of intelligent media asset management system,make collection-type database play a role to provide important resource support for media convergence development.This paper introduces how to use artificial intelligence technology to endow media assets with new vitality in aspects of resource storage,resource cataloging,scene segmentation and intelligent retrieval of traditional media asset management system,so as to enhance the role of traditional system in business production process.And through practical verification,it has reached advanced level of smart media asset management system in face recognition,speech recognition,OCR subtitle recognition and other aspects,and finally reached application requirements,which provides reference for business production and intelligent resource application of radio and television.