(Subscribe to our Today's Cache newsletter for a quick snapshot of top 5 tech stories. Click here to subscribe for free.)
Google’s AI team is working on a tool that will automatically convert a web page into a short video after the user specifies the duration and aspect ratio of the output video.
URL2Video extracts texts, images or videos along with design styles like font, colour, and layout from HTML sources.
While organizing the visual assets into a sequence of shots, it maintains a look-and-feel similar to source page. Using all these elements, it renders the material into a video.
To train the tool, the company interviewed designers who were familiar with web designs and video ad creation to understand important aspects of a webpage derived from heuristics.
The designer-informed heuristics capture common video-editing styles, content hierarchy, constraining the amount of information in a shot and its time duration.
URL2Video uses this information, analyses the content, and selects the video or picture while preserving the design styles. Then based on the user preference, it selects other elements and ranks key elements to churn out a video.
Even after the video is prepared, a user can change colours and style, and regenerate it according to their needs.
The AI team demonstrated the performance of the end-to-end URL2Video pipeline on a variety of webpages and uploaded the video on its blog. It showed that the tool automatically makes editing decisions on font, colour choices, timing and content ordering in a video.
The company is also developing new techniques that support the audio track and a voice-over in video editing.
“We envision a future where creators focus on making high-level decisions and an ML model interactively suggests detailed temporal and graphical edits for a final video creation on multiple platforms,” Google said in a blog post.
