
MemNet can actually tell you how memorable is your Facebook profile picture.
If taking selfies has become a thing now why not go to wave and make the experience as epic as possible? Now you can do just that, thanks to some rather brilliant work from the boys working for MIT’s Computer Sciences and Artificial Intelligence Center. The group who came up with this new gizmo asked themselves the following question: how memorable are your selfies?
In many instances, the answer to this question would depend on several factors, much of them being related to how the said photo makes us feel or rather what feeling does it stir in us. Many would think that computers and feeling don’t go well together and that the matters of the soul are not for a tin heart.
But it would seem that the research team for MIT might have found a way around this problem regarding robots with feeling. Unfortunately, they didn’t come up with a robot that can quote Shakespeare from the heart, but they’ve managed to develop a software which is capable of showing you how cool your Facebook photos are.
The computer program uses a newly-devised algorithm called MemNet. The algorithm makes use of the latest research in the area of computational photography. This new and bold domain uses several deep-learning algorithms in order to correct lens aberrations or to create other types of effects.
Now, regarding the new program developed by the research team, as we said it has an integrated deep-learning computer algorithm which is capable of analyzing your pictures. The program is designed to give you a rapid feedback regarding certain points of interests included in different photos.
Basically, the software is actually capable of telling you if your selfie is memorable or not. And, according to the team of scientists the algorithm has an uncanny accuracy when it comes down to saying “yes” or “no” to your pictures.
For now, the team is trying to enhance their algorithm, which could be included in various online applications. As stated, the software can serve many areas of study. Moreover, the algorithm can even serve an entertainment purpose.
The developers suggested that the algorithm might we well suited for developing applications used in mass media, marketing, sales and even in movies.
But for now, the algorithm is still being tested in order to rule out any possible bugs which may impede its snappy feedback.
Although the deep-learning gadget may be years away from making its appearance on the big state, we are still confident that it will be an earth-shaker and a heart-breaker.
Photo credits:www.flickr.com