Have you ever seen a dress in a store and wanted to know the lowest price you could get it ?
In this AI project, you can build an app that allows users to upload a picture of the item they want to buy. Then, the app will scan through many online stores and find the lowest price for the item . This way, the user gets the best possible deal.
To create an app like this, you will first need to create an algorithm that can identify objects in an image. For example, if the user uploads a picture of a pink floral dress, the algorithm should identify the colour and style of the dress correctly.
You can use transfer learning for this AI project and train on top of models like VGG-16 with a pre-existing database of item descriptions. Once the model is built, you can give the user a choice to specify additional information about the item — brand, outlet, etc.
After collecting all this information, you need to build an algorithm that identifies online stores based on the brand information provided. Create an automated tool that opens these sites and scrapes pricing information from at least 3–4 online stores.
Then, return the site name and pricing information to the user, along with a link to where they can buy the item from. The only part of this project that incorporates AI is the item description based on the image uploaded by the user. Everything else requires you to have model deployment skills, the ability to render information quickly to the user, and a firm grasp of data science programming languages.
Dataset: Kaggle Fashion Dataset
