Recently, Google researchers introduced a new AI system that converts a web page into a short video, known as URL2Video. According to the researchers, given temporal and visual constraint, URL2Video is an automatic approach that converts a web page into a short video. The AI system captures quality materials and design styles extracted from a web page, including fonts, colours, and layouts.
Video production entails a sequence of choices, such as the content that is best suited to a target audience, the position to available the assets within the field of view, the temporal arrangement that will yield the most compelling narrative and so on. While editing one single video is a time-consuming process, producing multiple videos from the same theme can be even more challenging.
To mitigate such issues, researchers at Google have been exploring this field by developing machine learning and computational models that can help create multimedia content in a faster and efficient manner. With the help of this AI system, one can now leverage the existing online assets, such as a website to create unique videos in a short amount of time.
Behind the Model
URL2Video is an AI-based video-creation tool automatically converts a web page into a short video, given temporal and visual constraints provided by the content owner. The researchers stated, “URL2Video builds on prior work of design understanding and computational techniques for video creation.”
They added, “Our work is built upon existing techniques of web and ad understanding, but we focus on converting web content to a short video given both the temporal and spatial constraints with design principles.”
The system extracts assets, such as text, images, or videos as well as their design styles including the fonts, graphical layouts, hierarchy, etc. from HTML sources and organises the visual assets into a sequence of shots while maintaining a look-and-feel similar to the source page.
Also, the creators can review the video composition, modify constraints, and generate video variation through a user interface. The design engine of URL2Video organises the materials and creates a sequence of shots to fulfil the temporal and spatial constraints. Also, in order to enable creators to iterate the video variation quickly, a user interface is provided that visualises the selected assets from the web page and its storyboard.
How It Works
When a user provides an URL to a web page that illustrates the user’s specific business, according to its developers, the URL2Video pipeline automatically selects key content from the page and decides the temporal as well as the visual presentation of each asset.
Stay ConnectedGet the latest updates and relevant offers by sharing your email.
The assets are based on a collection of heuristics that are derived from an interview study with various designers who were related to web design as well as video ad creation techniques. Using this information, the pipeline of URL2Video parses a web page, analyses the content as well as selects visually salient text or images while preserving their design styles, which it organises according to the video specifications provided by the user.
Advantages of URL2Video
Some of the advantages of this AI system are mentioned below:
- URL2Video can effectively extract the design elements from a web page and support designers by bootstrapping the video creation process.
- URL2Video captures quality materials and design styles extracted from a web page.
- This AI-based video creation tool can reduce the time as well as efforts that are taken during a video creation process.
- It is an end-to-end solution that renders a web page, extracts its assets and styles, and automatically makes both temporal and visual decisions to composing content in a video given user-specified constraints.
- URL2Video can produce quick marketing videos for small businesses, and it requires no video editing experiences.
You can check the demonstration of URL2Video here:
If you loved this story, do join our Telegram Community.
Also, you can write for us and be one of the 500+ experts who have contributed stories at AIM. Share your nominations here.
What's Your Reaction?
A Technical Journalist who loves writing about Machine Learning and Artificial Intelligence. A lover of music, writing and learning something out of the box. Contact: email@example.com