https://docs.google.com/document/d/1OFYEgNUBwMRJavC272gSgiO8l5eE0ctqDjbB3z7gzSw/edit?usp=sharing
By Ryan Teoh
Many people have issues finding what they want to buy, and how they can get the best price and quality product. You have to scour the web, look on thousands of different reviews and stores, decide which item to buy, and where to buy it from.
To solve this problem, I built BargainWorthy. All users have to do is upload an image of an item to the website, and using machine learning and the Amazon Data Service API, the app returns similar products the user might be interested in buying, along with information about each product.
Flowchart of web app
First, a user uploads their image of an item, for example a Cannondale mountain bike.
Then, the webpage sends the image to Azure Functions, which acts as the hub for HTTP requests.
Using a GET/POST request, Azure Functions then sends the image to Microsoft's Computer Vision API, a machine learning algorithm that scans the image and returns the objects and brands detected.
These objects and brands are then used as keywords, and sent to the Amazon Data Service API, which return various product recommendations the user may be interested in.
Now that we understand how it works, let's try coding it together!
First we'll work on the Azure Functions HTTP request, and then we'll code the webpage that calls that HTTP request correctly.
In order to replicate this project, you will need: