Skip to content

Latest commit

 

History

History
8 lines (5 loc) · 913 Bytes

README.md

File metadata and controls

8 lines (5 loc) · 913 Bytes

Eye-PERS

  • Improved Personalized Eyewear Recommendation System. (eye-PERS) Our project is a recommendation engine for eyewear on Lenskart. It aims to provide an enhanced and personalized experience for users.
  • Technologies used: Object Detection, AR Kit, Caffe, HaarCascades, Image Processing - Open CV, AWS rekognition: (facial detection and analysis API), AWS personalized API, Canny Edge Detection.

The problem eye-PERS solves

We are solving the ambiguity problem in choosing the frames while ordering spectacles. Generally, with popular eyewear-selling companies like Lenskart, we don't get personal recommendations. The recommendation is only based on the site's best sellers, which may not be an excellent fit for a particular person. We solve this problem by analysing the person's facial features and using these features and artificial intelligence to recommend what's best for a given person.