Google’s new artificial intelligence tool, URL2 Video automatically converts webpages into short videos.Leading tech giant recently talked about its new AI in a Google AI blog entitled, Experimenting with Automatic Video Creation from a Web Page. Blog post said that using machine learning by utilising assets from the existing webpages can give creators a head start in the process of video making.
It noted, “Businesses commonly host websites that contain rich visual representations about their services or products, all of which could be repurposed for other multimedia formats, such as videos, potentially enabling those without extensive resources the ability to reach a broader audience.”
Often a usual video may take businesses up to a few days or weeks and may cost up to $1,200. But, with the help of AI and ML, new Google tool can play a pivotal role in reducing both time and costs of the process. Dedicated Google AI systemwas first presented at the 2020 User Interface Software and Technology Symposium.
How URL2 Video change source webpages into short video?
The new AI systems operates by extracting content including text and images from the specific webpage and their design style covering colour, font and graphical layouts from HTML sources. Subsequently, it organizes all the assets into a sequence of shots in a look and feel similar to the source page.
The consequent videos and presentations come from a set of heuristics which were identified earlier through a study with designers which captures various video editing styles such as content hierarchy, constraining the amount of total information in a shot and its time duration and also takes into branding factors like consistent design style and colour combinations.
URL2Video then utilise this knowledge to select text and images and preserve a webpage’s design style. GoogleAI system extract all the information and multimedia material on a per-webpage basis and also identify visually distinguishable elements like headings, product images, descriptions and call-to-action buttons in the process.
Notably, it covers both the raw assets as well as detailed design specifications through HTML tags, CSS styles and more for each element. In the next step, URL2Video assign all assets a priority score on the basis of their annotations and visual appearance and ranks them. Notably, an asset group that occupies a larger part of the source webpage’s top section also receives a higher score.
After automatically optimizing the total priority score, AI system presents only dominant elements like a headline and few of the multimedia assets. Once it has completed the priority list,it follows the heuristics obtained from the aforementioned design study. It lays the foundation for colour combinations as well as temporal and spatial arrangement.
In the last phase, AIO tool transfers the layout of elements into the video’s aspect ratio and apply all the style choices rendering the content into an MPEG-4 video. Additionally, Google evaluates these automatically-generated videos with its in-house team of designers for feedback.
So, what do you think of the new Google AI system converting webpages into short videos? Share with us in the comments below.