What is that mushroom ??!

Welcome on my new project ! The aim of this project is to recognize easily mushrooms that you pick when you are walking in the hood. You will just have to take a picture of a mushroom and the AI will automatically tell you what mushroom it is and if you can eat it or not.
Mushroom picking has always been part of the french culture and everybody wants to take part of this autumn tradition.

Introduction of the project

Indeed, when you're walking in the hood with your family or your girlfriend and you find a clearing with loads of mushrooms on it but you don't know what type it is because you don't really have knowledge on mushrooms. You won't put your health at risk to see if it is eatbale or not. No worries ! Thanks to " What's that mushroom ? " you'll directly know if you can go home with a bag full of mushrooms or not.
Mushroom picking has always been very selective because it requires some background but now everybody will be able to go and find its own mushrooms.

Mushrooms recognition

My project uses machine learning algorithms, through one of its classic applications : image classification. I'm using a pre-trained model called MobileNet, a machine learning model trained to recognize the content of certain images. Moreover, my project is based on ml5.js, a Javascript Library accessible on the web which provides access to machine learning algorithms and models.
How did I create this application ?
I used the "Getting Started" part of ml5 website. It was very useful to understand the functionning of ml5 and MobileNet.

1. First you have to take a picture your mushroom : 2. Then the AI will compare the image to his base and give you a result with a confidence on the result.

Label : Cepes
Confidence : 0.96

This step is done thanks to supervised learning : making prediction with a labelled dataset. Thanks to a base of images of mushrooms where it is specified each time which mushroom it is, the AI can make a prediction on the type of mushroom on the photo.
Everything is coded on the JS file, here is the JS part of the code with some explanations to code the app. I used the example here on the ml5 website, it is very well explained and you can easily understand how to make your own MobileNet website.

3. Then you can go back home with all your mushrooms !!!