Why should girls get all the likes? Statistics say that girls end up with more likes than guys. So we decided to make a machine learn what makes a picture popular and help guys get more likes by choosing the right picture!
What it does
Our machine learning algorithm helps you gauge the popularity of your image. You have to upload a potential profile picture and it will let you know if the picture will be popular or not.
How I built it
First, we crawled for images and associated likes from my friends list. Then we converted images to n[d] vector and used the radial vector machine to train and predict using the fed image data.
Challenges I ran into
Normalising the data set, reducing the data dimensions of the image data to enable it to run within given hours.
Accomplishments that I'm proud of
Tried implementing neural networks. P.S. Also, we found that number of fake profiles in FB(Its a whooping 83 Million- 1/10th of world population!!!!!!).
What I learned
Neural networks, machine learning, crawling.
What's next for Hot || Not
You would no longer have to waste so much time in choosing a popular profile picture. The algorithm will pro-actively suggest what image to use. It is also useful to photographers to choose the best photographs.