Tedious Film Editing Could Soon Become a Thing of the Past

For those of you who have tried to edit your own film or video, you will know how painstaking it can be sifting through hours and hours of material just to get those one or two crucial minutes you need. It’s a long, drawn out process, but one that is necessary for final production. Thankfully a team of Stanford computer scientists has developed an editing tool that can speed up this process no end. Their editing tool can make a rough cut in a matter of minutes opposed to hours, leaving them more time free to get on with the more creative tasks. The team is intending to showcase their masterpiece technology at the upcoming SIGGRAPH conference.

Most traditional film editing techniques involve the editors having to watch and re-watch the same footage over and over to find the most suitable camera angle and performance for every line of dialog. Eliminating some of this will save editors an enormous amount of time. With the new tool, the editor would just have to feed in the raw video to the software along with lines of script that relate to that scene. Then, the software will automatically process the footage into short clips, labeling them using various image and language processing technologies such as face-tracking to identify who is speaking. It can also detect whether the scene is a close-up, medium, or wide-angle view shot. Combining all of this the software then strings all the clips together to form a rough cut of the footage.

The most challenging part of making the tool was turning the most common idioms of film editing into mathematical constraints that the software could understand and implement.  With these idioms understood the software looks for all possible sequences of clips and uses its artificial intelligence (AI) techniques to find a path through the clips that conform to all the requirements set out by the idioms. Team leader of the project and a professor of computer science, Maneesh Agrawala, commented, “It will allow editors and storytellers to focus on the important parts of the job – figuring out what sequence of cuts will tell the story the best.”

While the current version will only work with dialog scenes, the researchers are hoping to develop the tool further to work with other story types too such as action or how-to-videos. Agrawala also hopes that eventually the tool will be used during the live filming for even more time-saving. “It’s still going to require a professional editor to produce a professional-quality edit,” states Agrawala. “But, it should make their job much easier and allow them to be more creative and explore the design space of possible edits much more easily.

Article Via Standford University / Engineering 

More News to Read