Smart-Wardrobe-Planner

An app that matches clothes and recommends the user best combinations taking into consideration user's attributes

Clothes Classification:

• Done using CNNs into 5 classes-tshirts, shirts, casual shorts, pants, jackets
• Dataset link:https://www.kaggle.com/paramaggarwal/fashion-product-images-dataset

Clothes Matching:

• Create Dataset of upper body clothing matched with selected pants -->Done
• Train a binary classifier for each of the pants thats extracts features of matching shirts/tshirts and finds if it is matching with the given pants
• When user inputs new pant, we find the closest of the selected pant using image embedding and check if the shirt/tshirt is matching with the selected pant using the corressponding classifier

Occassion Based Recommendation:

• Select shirts/tshirts and pants that satisfy the occassion use the clothes matching algorithm to recommend clothes. This satisfies the wear properly and aesthetically criteria