PicSift is a simple to use photo aesthetics system that helps people quickly find their best photos. Using commonly accepted rules of photography and image statistics, the application extracts relevant features to determine if a photo is good. Training and test data were webscraped from multiple sites using regular expressions. The application back-end was written in python with the help of OpenCV, SciKit Learn/Image, Pandas, NumPy and SciPy, while the front-end used Flask and Bootstrap.
kumbhani/PicSift
Folders and files
| Name | Name | Last commit date | ||
|---|---|---|---|---|