At Polarr, we strive to turn complex thoughts into real world and beautifully designed experiences. Internally, we have a thriving and unique research team providing an assortment of machine learning models that are critical in assisting our user’s creative intention. This complex thought, from intelligent cropping, intelligent aesthetic ranking, intelligent filtering, to many other use cases we’re envisioning, supports the intricate processes that drives innovation at Polarr.

We often ask ourselves, "how do we package our technology into a product that drives a positive user experience previously thought not possible?".

Memoir is an unusual cousin to Polarr's current product line, an experiment to creatively combine A.I. modules at low overhead to instill a new lively persona within Polarr.

The precise goal we wanted was to create an experience that welcomes users to be happily surprised at unearthing old memories they have forgotten. The main engineering challenge stems from the low consistency of quality for old assets viewed. Do not get us wrong, these photos still hold value, but the journey to revisit them hasn’t seen enough innovation. We want the experience of revisiting memories be rewarded and incentivized through authentic means.

To accomplish this goal, we decided that intelligent aesthetic ranking, an in-house Polarr A.I. module would be a key part to allow Memoir to pick interesting photos from the past. This model aesthetically ranks photos at remarkable speeds. It can be used against randomly picked photos or bounded by a timeline of your choosing to single out the best ones. The reason is simple, these photos mean a little more to the user than the rest.

If generating interesting photos from old memories succeeded in reminding you to use Memoir, how can we sustain this feeling of remembrance each time? We made a conscious decision that re-openinng Memior should be a new experience every time, and the timeline will never look the same as the first time you used the app, so that the users can seize this opportunity to re-experience a different set of memory, and also to do more. But what more could we do?

Our team looked at different methods to produce a meaningful creation out of picked photos, and we agreed on the relative sequence of complimentary memories stitched together is vital to creations of Memior - Live wallpapers and collages are two common mediums that fulfilled this goal. To create a live wallpaper in Memoir, a user could find a sequence they like, adding as many memories and removing as many as they want, and then generate a memorabilia that is always available via a gesture from the lock-screen. The collage is what it is, a collage, and it is less heavy, but a higher quality export, meant for easy sharing and safe keeping.

Throughout the content generation process, Polarr’s intelligent cropping, another in-house A.I. model, assists in the way Memior packages images into videos and collages, focusing on the important details of an important image. The aspect ratio of a square or the screen of a device guides the model to find the best angle that compliments the image, preparing collages or live wallpapers to be shared.

For the techies and other developers working at Polarr, Memoir focuses on delivering a taste of Artificial Intelligence on the edge, through Polarr’s brutal eye and simple tongue. The simple things we miss, the angles we could have seen. This is one of the many reasons I am grateful to be a Polarrian, to put technology like this and what is to come in your pocket, no matter the cost or affordability of creating. We strive to provide the power of the professional into the hands of the amateur. Breaking the expensive barriers of expression and bringing it back to the roots. Just you.

There is much more to come this year from us and many more projects to share assisting people around the world on their adventure of expression. Make sure to stay tuned and to follow us on our journeys as well. We hope you will enjoy Polarr Memoir as much as we do.

Download Polarr Memoir for iPhone and iPad