Open Source Subnet An independent Open Source community View more

Facebook Hyperlapse: Your friends' videos, but better

082614 facebook hyperlapse

Hyperlapse merges image stabilization and fast forward to automatically compress mobile videos into short stories, sidestepping Facebook video fatigue.

Face it -- most mobile videos shot by your friends are hard to watch. The popularity of Vine’s seven-second videos has made this abundantly clear. Unless it’s a video of a cat attacking a dog to protect a toddler, longer videos are just not that engaging and won’t be shared that much.

As good as the cameras are in the new smartphones, after just a few seconds the videos get the jitters because smartphone cameras can’t be held perfectly still. And unless you’re a parent or a grandfather watching your children, smartphone videos are mostly too long for almost anyone to sit through.

Hyperlapse gives Facebook a differentiated product to compete with Vine by merging a couple of very well-understood technologies: image stabilization and fast forward time-lapsed video. It filters out the jitters and speeds up the video playback to avoid the boring disappointment that most viewers feel shortly after clicking on a friend’s Facebook video. Condensing longer videos, Hyperlapse has a Vine-like quality. Vine videos are very approachable because the video either makes a point quickly or mercifully ends quickly before the pain of boredom sets in. No guilty feelings from closing a friend’s 22-minute video of his or her toddler 15 seconds into it.

Image-stabilized videos aren’t new. The LG G2 announced a year ago incorporated optical image stabilization. YouTube has image stabilization too, but it is a two-step process: upload the video to YouTube, and then run the stabilizer.

Google Plus Stories was developed along the same lines, but it is designed for still images. Because images are tagged with geographical location, Google Plus Stories knows where the pictures were taken. The images are first run through Auto Enhance, a Google cloud-based application that algorithmically improves the images. Then some of the images with algorithmically high scores are incorporated into a Flipboard-like sharable presentation, including maps showing the journey. Google Plus does it automatically and serves up a link to share.

Google and Facebook solve the problem by automatically turning videos and images into shorter, more compelling and easily digestible stories. With consumers’ short, 140-character attention spans, it’s too hard to sit through more than a few seconds of friend’s shared amateur mobile video or click past the first five posted images, especially when a whole Facebook or Twitter newsfeed awaits.

Hyperlapse isn’t a technical breakthrough by any means; but it is really smart applied engineering. Facebook’s goal is to keep people engaged with user-generated content so it can increase value of advertisements in the newsfeed. If users pass up their friends’ videos or move on to the next post out of boredom, the overall experience is devalued and might make them less likely to open Facebook. Short, fast-motion time-lapsed video is a good solution to Facebook video fatigue.

Fast-motion time-lapsed videos are a good point of entry. But to keep users’ attention, Hyperlapse will need algorithmic intelligence -- a smart editor per se -- that chooses the best parts of the video, similarly to how Google Plus Stories selects the best pictures to create a compelling normal-speed video with the brevity and impact of time-constrained videos.

Insider Shootout: Best security tools for small business
Join the discussion
Be the first to comment on this article. Our Commenting Policies