On February 29, 2020, the International Federation of the Periodical Press Ltd (FIPP) launched the FIPP Insider to cover news, insights and perspectives from the media.
“An analysis of key media trends in the press and publishing industry over the past 10 years shows that time and depth of application of new technologies are important drivers,” said James Hewes, CEO of FIPP. “While the subscription model has always been one of the main strengths of magazines, I think the majority of revenues will be generated by membership and digital subscriptions for the next 10 years, and it’s time for the publishing industry to go videolized.”
Podcasts and other forms of audio and video content have become increasingly popular. In January 2020, the British newspaper The Times announced that it would launch its talk show radio station later this year. Existing shows such as The Media Podcast of Olly Mann, Media Show of BBC and Media Voices Podcast are widely well received.
A critical period for Press and publishing enterprises to apply intelligent image technology
Although it is well known that the decline of print advertising and decreasing circulation herald the end of the printing era, the trend is beyond doubt that media content creates information value through multiple forms, carriers and platforms.
In November 2019, Forrester, a market research company, released 2020 Content Payment, predicting that the podcast will become the next $1 billion media market, “which will be a very significant growth,” Forrester Mediapost pointed out. According to The Podcast Consumer 2019 published by Edison Research, Vlog’s share of audio listening has more than doubled in the past five years (increased by 122% since 2014). Overall, about 90 million Americans, or just under a third of Americans over the age of 12, listen to podcasts every month.
The maturity of application algorithms of Chinese AI companies enables most publishers to upgrade their products to the form of “text plus audio-visual”. For example, when reporting breaking news, Caixin often publishes audio-visual and then deep plus visual graphic content, making it one of its paid products.
Statistics from Douyin, the Chinese version of Tik Tok, show that as of December 2018, 1, 334 traditional media have started to use Douyin (AI), which is attributed to the maturity of AI video technologies and the popularity of Automatic Generated Content (AGC) in the publishing and media industry. In fact, typical application scenarios of intelligent image technology in China include news short video (1's-5's), Podcast (5's-30's), news visualization and virtual anchor.
Media integration is a technological revolution that requires more thinking on the application of technologies in industries
It is precisely because AI technology improves production efficiency and quality that media and entertainment companies attempt to make “smart video” part of the mainstream content. According to Li Baoshan, Director of the People’s Daily, “Media integration is not only a transformation brought about by the technological revolution, but also a profound change planned and advanced at the national level.”
Han Yi, Deputy General Manager of the Technology Business Division of CCTV International Network Co., Ltd. said during a public event, “Take the World Cup for example, we make visual optimization analysis and optimization effect evaluation of each game video in aspects of high-speed sports scenes, football movement tracking model, capture of the audience’s expressions, etc. Compared with the traditional manual mode, the use of AI technology, while ensuring the quality of user experience, enables the system to start the automatic alarm mode through automatic service adjustment to inform the operation and maintenance personnel to intervene when encountering a failure threshold.”
Beijing Moviebook Technology Co., Ltd., an AI technology company, said that it makes use of the three-dimensional convolutional neural network based on machine learning to analyze and discriminate highlights in videos (such as dunking, blocking, snatching, wonderful three points and points gained through excellent cooperation in a basketball match). This together with basketball star recognition helps position star’s highlights precisely and thereby forming a collection of players’ wonderful moments in a game. Based on AI image recognition technology, Moviebook Technology introduces a feature-based recommendation system framework. It emphasizes the importance of news value, integrates news public opinion guidance, social effects, etc. into the algorithm and modelizes news value. By adopting this framework, the media can reconcile news value with user preference when spreading news and distribute the two collaboratively on platforms.
In practice, compared with traditional manual video editing, the use of AI can reduce the cost of video production to three thousandths and increase the speed of video editing by 40 times. It has not only greatly improved the visual experience of the audience, but also quickened the production of collections of video highlights in sports events and sharply cut production costs.
An analyst from Industrial Securities Co., Ltd., who specializes in media integration, contends that it is not new for China’s media and entertainment industries to capitalize on AI technology to improve production efficiency, and that the technology integration plan has been adopted by most media, such as People’s Network, CCTV Network, Xinhua Network and Guangming Network, in completing their respective media integration.
“The first task for AI technology teams is to apply technologies to industries. Technology companies should first win media users’ trust and meet the end audience’s requirements for image quality. This requires AI technology providers to not only strengthen its technologies, but also accumulate operation experience and gain a sound understanding of the upstream and downstream of the industrial chain.” The independent technical analyst continued to stress that “Moviebook Technology has a deeper understanding of ‘how technologies can be made a good use of to become an important part of the industry’, which is crucial for technology companies”.
Perhaps, the large-scale application of AI intelligent image technology marks just the beginning of the combination of culture, media, entertainment industries and technologies. AI will be adapted to many other similar scenes, applied in a wider and more in-depth fashion, and give rise to more cutting-edge innovations.