Pantomath

Inspiration

We figured out that one of the major problem customer face in online shopping these days searching for an item without looking up for the specifics of the product. So rather than searching the product manually user can upload the image of the product he is looking for.

What it does

This web-app takes a image an input and it will show user the most similar products for that image as an output. For example if you have a pic of Shoe and you want to buy same kind of shoes so you can directly upload that image and it will show you all the similar shoes ...that you have uploaded

How we built it

We build a convolution neural network with the weights of VGG16 model. This convolution neural network takes a single image that user want to search as an input and it will return top 5 products those has highest match with input image. We also use VGG 16 model to get feature's from an image so we can use those feature in search query to make result or recommendation more effecienct. technologuies

Challenges we ran into

The biggest challenge that we faced is to integration backend (python) code with UI. So we build convolution neural network on colab(because of GPU) and to use our model as an API we use a FASTAPI But while implementing FASTAPI with UI code we face lot of challenges. But finally somehow, we made it.

Accomplishments that we're proud of

We can confidently say that we tried to achieve the given task in a short period of time.

What we learned

We functioned effectually as a team hence thereby comprehending team management skills and grasping the abilility to work under pressure productively.

What's next for Pantomath

We will build fashion prediction model which will help suggest the user a fashion style according to his personal preferences as well as the trending fashion.