Imagion is a UCB Data Science program capstone project. Imagion helps you pick which picture to upload to Instagram. Whether you are an advertiser that needs help deciding which creative will get the best click-through rate or an influencer that wants to maximize likes, Imagion is here to help.
Our team used machine learning to train a convolutional neural network to rank input images on a scale from 0 to 10 based on how well the model believes the images will do on Instagram. You can upload anywhere from 1 to 5 images at once and our algorithm will provide near real-time rankings of images and their respective scores. Our model was trained on a dataset of ~12k images from Instagram.
- Sue(Xue) Yang
- Haroon Choudery
- Roiana Reid
Imagion Project on UC Berkeley Ischool Page
Home Page
About Page
Upload Page
Result Page
Contact Page
