21 Feb, 2019 – Forrester has released its 2019-2020 predictions identifying the major dynamics of AI Technology that will affect companies in the coming year – two years where Artificial Intelligence will shift from strategic ambitions to pragmatic, application solutions. So, VideoAI will first industry to using is Entertainment Media.
In 2018, Entertainment Media leaders set their sights on large-scale initiatives such as digital transformation and Auto experience (AX).
However, many faced the harsh reality that these strategies are hard, costly, and challenge the way that leaders run their businesses. The pace of change is not slowing down, nor is the market getting more forgiving – Customers challenge companies to deliver value to their lives. However, it is undeniable that Artificial Intelligence must, or has brought about significant changes in human life, the convenience and change.
“The pace and diversity of technology, from the promise of AI to the reality of technical debt, challenges prioritisation and budgets,” says Forrester chief research officer Carrie Johnson in a blog post.
This is why in 2019-2020, Entertainment Media leaders will turn their attention to pragmatic, application solutions.
For Video technology practitioners, Forrester predicts that:
1、Artificial Intelligence brings revolution to Image Analysis Technology. The comprehensive application of Artificial Intelligence has brought big imagination to the video industry. In particular, the understanding and reconstruction of dynamic visual content, combined with computer vision and animation, will produce new Intelligent Video Technology, Maybe before others is the Auto-Generate Video Technology.
2、VideoAI Technology Brings Strong Motivation for Industrial Change. The Entertainment Media Industry has become an important pillar of the development of the digital economy. The Digital Entertainment Media Industry represented by the content industry has shown a spurt of development in the past few years. However, video conten and video platform continuing to grow are facing real challenges. Leaders improve key productivity by AI technology and innovation application in products and services , this goal has been verified in a wide range.
3、In 2019-2020，the key Revenue growth will coming from VideoAI Technology applications in multiple scenarios and multiple dimensions. VideoAI Technology becomes the core-engine for realizing the commercial value of Entertainment Media industry, most importantly, Revenue growth in the fields of Auto-Generate Video, Ad-Marketing Video, Video Deep learning, Big data Video, Video-to-Data, Video RPA,etc., through automated image processing and production. Technology powers Entertainment Media industrial upgrading.
4、VideoAI Technology will expand the business ecology. AI companies need to focus on industry practices, closed-loop business models, and open platform technologies to promote the coordinated development and further prosperity of the intelligent image ecosystem.
Forrester Report： VideoAI Technologies And Solutions
Application Scenes of Video AI Technology in Industry
The global entertainment industry will exceed $2.6 trillion by 2021, while the penetration rate of VideoAI technology in the global entertainment market will rise from less than one in ten thousand to six in one thousand. “The potential huge value has been shown of using AI and computer vision technology, batch production of Auto-Generate video will formed a trillion-scale market,” said a VP of AI company.
For VideoAI technology applications of scene, Forrester predicts that:
1、 Content Creativity and Auto-Generate video Making.
According to Forrester interview, VideoAI technology solves the problem of large-scale video content production and high cost in Special effects production. Using VideoAI technology platform, it can improve the video content realization rate and production efficiency, and reduce the labor cost by more than 60%. “Finding Nemo 2” and “Finding Nemo 3” used VideoAI technology of the first time, by videoAI technology, defects in the film can be automatically corrected in the production process. Artificial Intelligence and Deep learning are changing and reshaping the process from creation, design, shooting and production to post-effects in entertainment media industry.
In response, an internationally renowned director of investment said, “In 2017, the global box office reached 40.6 billion US dollars. VieoAI is influencing the entertainment industry and gradually changing the production process of film and television content. The Auto-Generate video era is coming.”
Forrester Report：VideoAI Technologies And Solutions
1、 Video to Data Structuring Based on Content Intelligence.
With the improvement of AI’s correct rate of content understanding, it can describe video digitally, which is the most solid foundation for personalized video recommendation in the later stage. Video frames are decomposed by video structuring technology, and then the frames are classified. Then the video and frames are labeled to understand which category the video belongs to. On this basis, it can realize massive video classification, review, reproduction, intelligent retrieval and personalized recommendation applications.
Video recognition and analysis will take the lead in large-scale applications. By understanding the video, the best key frames or clips will be automatically selected as the video cover, highlighting the core content of the frequency, thus attracting users to convert to watch.
Video to Data structuring helps to make a substantial leap in video content auditing to solve the long-standing problem of video content auditing that has plagued video operators or channels. In the next two days, it will be automatically completed through API interface or automatic triggering of workflow.
Video structuring will play a more important role in the copyright industry. In addition to playing a key role in distribution and consumption, intelligent media search engine based on video fingerprint (unique identification based on multi-modal features of video) will not change with the key role of media file format conversion, editing, compression and rotation, and is widely used in video copyright protection.
2、 Intelligent Recommendation Enhances Consumer Digital Experience
Through intelligent cataloguing based on image content recognition and processing, more intelligent recommendation can be achieved according to user’s viewing behavior and content preferences, and the viewing experience of customers can be improved.
3、 Intelligent Video Editing and Broadcasting
As an example, Taking one hour video processing , traditional video editing takes several hours for professionals to complete. Through intelligent image technology, we can get rid of the high threshold and high cost of professional equipment, professional editing software and professionals, and reduce the cost by more than 100 times. For example, aiming at a large number of wonderful scenes of variety shows, intelligent image technology is used to analyze multi-dimensional panoramic images of characters, gestures, movements and trajectories, and to automatically edit and synthesize them. It takes only ten to tens of seconds at the fastest, and the production efficiency is increased by more than 10 times. Through multi-modal content understanding technology, the information of people and objects can be automatically superimposed in single/multi-channel videos, and the collection content can be produced intelligently in real time and accurately, and the multi-scene and multi-site Cooperative Intelligent broadcasting can be realized.
Forrester Report：VideoAI Technologies And Solutions
4、 Marketing Embraced Artificial Intelligence To Transform Its Business Model.
As a new way of content marketing, native video advertising based on intelligent image technology is widely used. Capture hot content while automatically generating new video scenes. Intelligent image technology is used to provide a large number of digital resources for downstream tasks, such as original video content marketing, game production, virtual online trading stores, and to establish intelligent IP operation monitoring model, value prediction and accurate promotion, including stars and literary works.
CTO of a well-known online mobile company surveyed by Forrester, also said, “AI’s application in video field is very direct, especially through automation technology to replace simple and repetitive manual, marketing field will directly cross the domain to the era of intelligence”.
Case Study ：Mivebook’s Plan To Auto-Generate video with Artificial Intelligence in70%
Experimentation And Business applicationAre Key To Thriving On Emerging Technologies。
Moviebook, China’s leading provider of VideoAI technology and solutions, is covering all hardware terminals and computer chips through the visual scene through the integration of video channels to achieve closed-loop intelligent image realization. Its intelligent image production technology has the characteristics of batch automatic processing, sub-pixel level analysis, intelligent superposition and traceless display.
Moviebook releases Motion Capture from Video System (MCVS) framework and products that can quickly realize image content production through Vedio AI content production engine. MCVS does not require pre-action capture to synthesize highly structured data. It allows machines to directly imitate a large number of existing video clips to learn difficult skills, allowing data-driven imitation to generate unsupervised learning video content. For example, capturing actions automatically from the video to learn, so that the machine can automatically regenerate a “brand new” video.
Native video marketing is a typical commercial application of Moviebook VideoAI: by extracting commercialized scenes in video, matching reasonable brand elements to add and replace content without changing the source, this way is conducive to deep mining of video dividends and achieving win-win situation.
Meanwhile, Entertainment Mmedia giants such as Disney, NETFLIX, HBO and TRONC are trying to replace standard content production with artificial intelligence. Forrester interviews found that the automatic production technology of video spectroscopy technology has been commercially used in the entertainment industry on a large scale: it can directly improve the image content liquidity, improve production efficiency, and reduce human costs by about 30%.
The first domestic music show “City of Miracle” launched by GBS has achieved remarkable results in the application of intelligent image production technology provided by music science and technology.
– Video scanning is analyzed by video neural network. Video is processed and marked according to video stream, frame rate, depth of field of shot, illumination, hue, scene, transit key frame and other factors. Video advertising incremental market is explored in batches.
– Based on the dimension and classification of video content, user behavior and advertisement attributes, the large data Deep learning algorithm automatically matches the most suitable scenes and advertisement strategies, and achieves the accurate delivery of the target population through intelligent optimization. Intelligently optimize the placement effect, so that the embedded advertisement and the color shape of the surrounding area can be combined with the video resolution to process the edge of the frame, so as to achieve the effect of native implantation.
According to Forrester, the “AI+” model has become an important support force for digital transformation and business innovation in various industries. Among them, VideoAI technology has been widely used in the entertainment industry. With the continuous development and maturity of this technology, it will inevitably enter more industrial application scenarios, which will enable industrial upgrading.