We all know that sinking feeling of opening a video that a friend has sent us because it’s “SO cute” or a “must-watch” and discovering that it’s five minutes long. (One of my relatives once sent me a 10-minute long music video that didn’t even have any words. I’ve got things to do here, people!) But in future you shouldn’t have to sit through any boring bits, thanks to Carnegie Mellon graduate student Bin Zhao’s new invention, LiveLight.
It’s an algorithm that scans videos for repetitive content and imagery and then cuts sections where nothing seems to be happening. It then stores these mind-numbing moments and compares them against other videos, so it can spot (and edit out) tediousness more quickly and effectively in future. It could make it much easier to analyse CCTV, but according to Wired, Zhao is keen for it to have consumer applications, too. This is going to be sad news for arty poser types and their languorous shots of empty landscapes, but good news for the rest of us. (Although I can’t deny that the idea of an algorithm to cut out the boring bits of things you read on the internet makes me a little nervous…)
Zhao apparently came up with the idea after seeing his mentor, Professor Eric P Xing, try to find the funny bits in videos of his son. The software was partly funded by Google and will be presented today at the Computer Vision and Pattern Recognition Conference in Ohio. You can see an example of how it works (and save yourself around a minute and a half) on YouTube.
In the meantime, if you already know where the fun part of a video starts and ends, you can use Tube Chop to manually select only that section. Tell your friends. And my family.
Image via Peter Taylor’s Flickr.