PÉPIN

Pépin

an iOS mobile app - a new way to discover flowers.

Role: Software Developer, UX Designer

Timeline: 2.5 weeks

Tools: Xcode, Swift, Machine Learning, Caffe Model, Wikipedia API

Team: Just me/Personal Project

Github Files: Pepin Project

—

OVERVIEW

Pépin, means “seed” in French. I created this personal project out of my love for flowers, and in order to practice using ML models. I wanted an app that could take a picture of flowers in my environment, and tell me what type of flower it is.

I often struggle with knowing how to care for different types of unique flowers I buy, and thus decided to create an application that provided this type of information as well.

I also wanted it to tell me more about the flower - fun facts about it, as well as easy body scrub recipes I could make using the flower (if applicable).

This is what Pépin does. Pépin, is an iOS application that allows you take pictures of flowers in your surrounding, and shows you interesting and useful information about it.

For the UX, I opted for a design that was clean and easy to navigate, while covering the core scenarios.

I developed the app in Xcode using Swift.

Identifying an image: For this task, I used the oxford102 caffe model that has been trained using thousands of flower images. I then converted this model to a .ML model file so I could use it in Xcode. Using the ML model, the app was able to identify the flower based on the photo of the flower taken.

Displaying Information About the Flower: To display more information about the flower, I made a HTTP GET requests to Wikipedia’s API using Alamofire and REST, to extract information from a selected flower’s Wikipedia page. I then rendered this information in the app.

As of now, the working app only displays the name of the flower, the Wikipedia photo of the flower, as well as the Scientific Data and Fun Facts. Please find my code files on GitHub, here.