Gul Ahmed has announced that a machine vision based visual analytics tool will be rolled out to outlets of Ideas by Gul Ahmed nationwide, tracking footfall, sentiments, and behavior to aid decision making.
Created by Integration Xperts, SenseR analyzes in-store CCTV systems to extract shopper activity data for retail analytics. According to Umair Azam, Managing Partner at Integration Xperts, SenseR uses facial expressions to record customer emotions by categorising facial expressions into the following broad categories: neutral, joy, sadness, angry.
“Proprietary deep learning cognitive algorithms are applied on CCTV videos which determine the existence of a face/faces in the frame and analyses each face for gender identification, emotion classification, and age approximation,” said Azam.
“For demographics, the system analyses face components which include but are not limited to, wrinkles, hair color, faical hair, and distribution of features on a human face. It is important to note that for gender, the system has been trained on local faces for a more accurate identification.”
With over 100 retail outlets nationwide, Gul Ahmed appointed Integration Xperts to help understand the customers’ journey in the store, identifying points in the store where the customers passed by or browsed.
According to Azam, this data will be co-related with point of sale data, stored in Microsoft Azure cluster to provide better insights. Utilising Blob storage by Microsoft, the system will store encrypted faces which are later used for inference to determine a returning customer, said Azam.
“SenseR provides two-way API integration which makes its data available for other systems to access and integrate with other systems” said Azam “The APIs are developed using Azure’s API Management tool allowing SenseR to create a scalable API gateway for securing, publishing, and analysing APIs and micro-services to internal and external consumers and applications.”
SenseR is a a retail demographics solution designed to augment existing tools and give them better insight about customer in-store behaviour, SenseR will be evaluated in its ability to help with sales forecasting, customer segmentation, time series analysis, and evaluating the impact of marketing campaigns on footfall generation.
