Summary

Create a binary image classifier to classify chakka and manga. You can get the dataset by scraping from google or you can use Unsplash api as mentioned in this article

Elements

• We need dataset containing images of mangoes and jackfruits
• We need machine learning model (Classic ML algorithms, CNNs, RNNs, whatever...)
• We need flask web app to make the project as an API (You can use fast api or the micro framework you prefer )
• Collect image dataset using Unsplash API
• Build the machine learning/ deep learning model in your favourite framework
• Train the model
• Write a small flask web app to host
• Test the AI model and deploy

Possible users journey

1. A user should be able to pass an image in a web interface or as GET/POST request with url as parameter.
2. A use gets the human readable classification as response - like "This is a mango" or "This is a jackfruit" or "Sorry, I don't recognise this fruit."

Acceptance criteria

• Its fine if the model is not performing that great!
• Should be able to send request to the API with image or image url as parameter and get classification result.
• Github repo with train, test and dataset preprocessing codes.
• Documentation

Helpful links