The emergence of eCommerce has disrupted the Revenue & Margins of the Shopping Malls & Retail Chains. They must compete with the eCommerce giants to survive.
The key challenge for the Malls & Retail Chains are how to attract more visitors, drive loyalty , conversions & improve the brand value. More the visitors, greater the probability to improve the Revenue (Shop Rental, Adv Revenue) and brand value.
The mall management does not have access to the Customers / Visitors data. So, to attract visitors without any demographics is a challenge. Another challenges is lack of accurate data on the current & historical footfall, the visitors profile, the visitor traversal path, dwell time etc.
With propel-R, we have solved this challenge using Deep Learning Algorithms. The algorithms are based on Computer Vision & Natural Language processing. With propel-R, we have solved the business issue of increasing visitor footfall, loyalty & brand value.
A shopping mall in India is looking for a visitor analytics to understand visitor count and demographic profile. This will enable the mall to maximize revenue from the rent, choose better mix of tenants, plan events for targeting right audience and put in place customer centric strategy.
Python, OpenCV, Tensorflow, Deep Learning
The solution involved building deep leaning model to detect persons coming through entry gate and people moving through corridors on different floors of the mall. The detected persons are then pass through age, gender and emotion model to derive these information. This is a cloud based solution hosted on AWS cloud for data processing. The edge device is used to capture the camera frames and post motion detection passed to AWS based server for further processing.