New Windows Phone app Photo Story uses machine learning to sort images

Image Credit: Microsoft

Even though the saying, “the best camera is the one with you” has been all but beaten to death, it’s perhaps more applicable today than ever before. Smartphone image captures are becoming the most illustrated form of photography around. Smartphones combine utility and convenience, which are making them the prefer choice for users who wanted capture every moment as quickly as possible.

Unfortunately, the point and shoot convenience of smartphone image capturing also leaves many users bulk dumping images into folders with reckless abandon. It isn’t until later on when users want to share images, that they run into the lackluster sorting options built into a smartphone’s operating systems.

Fortunately, most app developers have stepped in to fill the gaps with some creative solutions. A couple of years back, former Nokia devs created Lumia Storyteller for Windows Phone. The app serviced as an image aggregator that compiled metadata to create albums of photos.

Image Credit: Microsoft

Today, a Microsoft Garage Project called Photo Story takes the next step. Photo Story is an app that automatically chooses the best representative photos to create a themed video-album. Created albums come complete with editing tools, sharing features that include email and social media as well as stylized music applied automatically. Much of the selection process is done using machine learning. A similar form of image-based machine learning was shown off at Google I/O this year and WWDDC last year.

As machine learning is implemented more in sorting through big data, users will benefit from automatic contextual and procedurally learned presentations. According to Krishnan Ramnath, a senior research software development engineer at Microsoft, “Photo Story’s algorithm isn’t just looking for technical quality, such as whether or not it’s blurry or the exposure is correct. It also evaluates “interestingness.”

Image Credit: Microsoft

On top of Photo Story’s machine learning, Ramnath’s team also built in support for voice commands via Cortana. The user can simply say, “Story from today” to bring up auto selected photos from the day. Users will also notice a reduction in lag time between capturing and editing, “because Photo Story runs everything on the device, you don’t have to wait for uploads and downloads,” according to Ramnath.

Photo Story also offers users immersive editing controls that include transitions, background music, and stylized effects. Most importantly, Photo Story albums can be shared via email and social media. Historically, offerings such as Photo Story used proprietary software thus forcing owners and viewers into encouraging their friends and family into using the same app or using complicated workarounds.

Photo Story was released today on the Windows 10 Mobile/ Windows Phone Store

Source: New Windows Phone app Photo Story uses machine learning to sort images

Via: Google Alert for ML