Deep Learning: Photos to Video

Collectively, we, take a lot of photos. Managing and manipulating these photos takes a lot of work, and most people don’t do it.

At this year’s Google I/O conference, Google launched their new photo service, dubbed the “Gmail of photos“, alluding to the fact that this will “fix” photos much like Gmail helped improve our ability to deal with and manage email. Using simple, natural search, you can find that photos you’re looking for and they’ll automatically organize them for you. It looks amazing, and shows us where technology like computer vision and deep machine learning can take us.

Cornell University recently released DeepStereo that uses similar technology to do something amazing: turn series of photos into a smooth, continuous video:

This system takes multiple photos and uses the information in them to “figure out” what’s in-between. This example uses Google Street view photos as input, and generates videos from that:

With Google Photos, we’re starting to see where deep machine learning will take us – and looking at what’s in the labs right now, we can see how much farther this will go.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s