Business Problem

Currently, when the Yelp users submit a review, they can upload pictures alongside the reviews. 
In an era of photo-centric social storytelling, Yelp's users upload an enormous number of photos alongside their reviews.
However, when the users upload a review they need to manually select restaurant labels with appropriate categories such as good for lunch/dinner, serves alcohol, outdoor seating available, takes reservations etc. 
Since selecting the labels is optional, many users prefer not to select the labels which leaves many of the restaurants only partially categorized. 
The solution of this project is to build a model that reads a dataset of user submitted photos and automatically tags them with one or more labels relevant to the restaurant or business